Nyriad

nyriad

What Are the Top Startups in Oceania?

The Many Faces of Bitcoin

Hey Hackers! I’m Piyush Badkul and I’m a Research Engineer. First of all, a huge thank you to the HackerNoon community and staff for nominating me for a 2021 Noonies award! I have been nominated in the following categories please do check out these award pages and vote: HackerNoon Contributor of the Year - code HackerNoon Contributor of the Year - debugging HackerNoon Contributor of the Year - optimization As someone in the telecom industry, I believe that the most exciting technology of the present is 4G, 5G (and ongoing research in 6G) because telecommunication is the base of the society, and the better the foundation, the better the building. Better the telecommunication, better is the base and so is the growth of the society. Learn more about my thoughts and opinions on this, and my journey in the tech industry, via the interview below. 1. What do you do and why do you do it? I am working as a research engineer in the telecom technology center of the government of India where my work involves designing, deploying, scaling, and optimizing the core IP Multimedia subsystem (IMS) network and making it interoperable with multiple radio technologies including 3G, 4G, PSTN, etc. Basically, IMS is like a server for all the phone calls taking place through mobile phones and other devices. Telecom is the backbone of society. If you see it, if the network in your area fails, basically everything (digitally available) fails including transferring money or making emergency calls in critical situations. After working closely for a couple of years in the telecom sector, especially in India, I have seen the demands and challenges that lie ahead in this domain. Developing an indigenous solution for the above which can help numerous people of India keeps me pretty motivated and happy too. Apart from that, I believe the telecom sector in India has not fulfilled its promise and still has a  very long way to go. I want to be a part of it before it goes big. 2. Tell us more about the things you create / write / manage / build! The one thing that I would like to talk about, and I am in the process of building it too, is my Page . Basically, it is to help the people who are stuck in a situation similar to the one I was stuck in and to help the people gain recognition to the people who are working day and night irrespective of how successful (from society’s view) they are. It is the platform where I can share my thoughts about best practices in coding and keeping a positive approach in life. It is basically my way to contribute back to the community that has given so much to me and from where I have learned everything. CodesMyth 3. How did you end up on your current career path? Do you like it? When I was pursuing my bachelor's, my particular interest was actually Android and Data Mining. Among multiple projects in the above domain, I along with my peers launched an App on the play store named YoCount which would store all the tangible data (like memories) in the form of categories providing easy retrieval and sharing. But, during the campus placement, I was selected in my current company, where I was introduced to the delicacies of the Telecom domain. Not only do I like it, but I also love my current carrier path. 4. What tech are you most excited or passionate about right now and why? Right now, I am most excited and passionate about two things. One is 5G and the other is Blockchain. I don’t have perfect expertise, and I am just a beginner, but I can feel it in my Gut that the Blockchain is going to be the next big thing. Bitcoin is the first real solution to and is based on blockchain. Byzantine’ s general problem As of now, blockchain has been mainly used with respect to crypto and a financial aspect, but I would love to research the use of blockchain in the other domains as well, including the telecom domain. One very common use case appears to be storing the call detail records using blockchain technology, but still through research needs to be done. without a doubt, has been deployed in multiple countries, so there is nothing to explain about it. 5G, 5. What tech are you most worried about right now and why? As of now, generally speaking with respect to the web, app, or core development, I am not worried about the technology itself but am mostly worried about the way the technology is used. Machines with very high computing power are directed for attacks on security, digital frauds, bringing down economies and the wrong use of the dark web really concerns me. One such thing is the National security agencies that are tracking our day-to-day data with respect to our lifestyles, making privacy a complete joke. However, If proper research is introduced in technology, then there will always be a very bright future for that technology. Of course, there is a lifetime to everything, and every good thing comes to an end sooner or later. 6. If we gave you 10 million dollars to invest in something today, what would you invest in and why? If I were to get 10 million dollars to invest, my first choice would be definitely starting research in the blockchain. Of course, I don’t have it all figured out as to where, how, and why in blockchain, the money would be invested, but I do believe in the future of blockchain and would like to get a head start on it for sure. I have joined certain discord communities for blockchain discussions, but still, the way to use the blockchain apart from financial and crypto remain pretty unclear to me. 7. What are you currently learning? Currently, I am learning a blockchain course from Princeton University (available at Coursera) and EPC (Evolved packet core) network components. 8. What’s the best advice you’ve ever given someone? The best advice that I have given to someone is to start contributing to the Open-Source projects - this will involve multiple gains including proper documentation of codes, bug management, and version control management; long-distance peer collaboration, a sense of being part of something bigger, etc. Apart from that, whose work and programming language is similar to yours since there is always something extra that you can learn. I always insist others read the code of the open-source projects 9. What’s the best advice you’ve ever received? The best advice I believe that I have received is that always flow through a proper channel and don’t use shortcuts to make work easy. Going through the proper channels will always increase your knowledge, and as we all know: This game can be won through the power of knowledge. Apart from that, just be happy, because nothing good came out of being sad. Right? About HackerNoon’s 2021 Noonie Awards The annual Noonie Awards celebrate the best and brightest of the tech industry, bringing together all who are making the Internet and the world of tech what it is today. Please be sure to , nominate, and vote for the people and companies who you think are making the biggest impact on the tech industry today. check out our award categories The 2021 Noonies are sponsored by: , , and . Thank you so much to these sponsors who are helping us celebrate the accomplishments of all our nominees. bybit Dottech Domains Avast

The Myriad Benefits of Contributing to Open Source Projects with Piyush Badkul

2020 has been a hotbed of political and social debates. Curious, I got out my python script, scraped some academic data and set out to find what the Sociologists are talking about in September 2020. From coronavirus to lockdown, Black Lives Matters, US presidential elections, the politicisation of public health, trans rights, colonial guilt and climate change denial, the list is undeniably important and keeps growing. Trending in Sociology: Illustration by author. What follows are four topics based on the most popular papers that sociologists have been reading on the academic social network, Mendeley. If you want more details on how I collected and used the data, I’ve added a section at the end. Enjoy! 1. Race, Police Killings and Black Protest Strategies The risk of being killed by police in the United States Trending Paper: Risk of being killed by police use of force in the United States by age Researchers wanted to find out what was the risk of being killed by police in the United States based on your social group and age. Academics reviewed data from police-involved killings in the United States from 2013 to 2018 on a sample size of 100,000 people. What they found was very disturbing. The study showed black men were at the highest risk of death, with a 1 in 1000 chance of being killed by police in their lifetime. Peak risk occurred between the ages of 20 to 35 for all groups. This would mean, today, the leading cause of death for men of colour in the USA is secondary to police use of force. Because of its prevalence, the authors call for it to be declared a public health issue. The advantages and disadvantages of violent protest Trending Paper: Agenda Seeding- How 1960s Black Protests Moved Elites, Public Opinion An eye for an eye may be moral, but this research suggests it may not be strategic. Researchers looked to the recent history of the black-led protests between 1960 and 1972 to evaluate which tactics had the most impact, and which made things worse. Non-violent protests, it appears, were most effective when met with state or vigilante repression. It even had an impact on how people would votes in presidential elections, with the democrats seeing an increase of 1.6–2.5%. Protestor initiated violence; on the other hand, had a very different effect. It appears likely that the violent protests of the ’60s resulted in a 1.5–7.9% shift among whites towards Republicans. 2. COVID-19 and rethinking tourism — sustainability is not the answer Trending Papers: Post COVID-19 ecological and social reset- moving away from capitalist growth Socialising tourism for social and ecological justice after COVID-19 (2020) COVID-19, indigenous peoples and tourism: a view from New Zealand COVID-19- from temporary de-globalisation to a re-discovery of tourism Unsurprisingly, most of the worlds hospitality and tourism services have shut down due to COVID-19. Multiple trending papers have declared this as a rare opportunity to hit the reset button on the tourism and rethink how we can be more ‘responsible’. When we go on vacation, we are complicit in directly or indirectly taking advantage of a group of poorly paid and protected people. Many papers call for a strategy of ‘community-centred tourism’ which redefines and reorganises tourism based on the rights and interests of local communities/local people. Independently, sustainable tourism is not the answer. 3. Why changing your mind is hard and the influence of mass opinion The difficulty in changing our dispositions Trending Paper: Measuring Stability and Change in Personal Culture Using Panel Data (2020) When it comes to individuals changing their minds, it turns out there are two types of people: Those who actively update their beliefs when new information is presented, and those who never change their dispositions after a young age. Sadly, it seems the latter is more prevalent. After early socialisation, there is little evidence that most U.S. adults changed their beliefs or practices. The biggest accelerator for change? Generational turnover. Disappointing… Increasing polarisation in Society Trending Papers: Pluralistic Collapse- The “Oil Spill” Model of Mass Opinion Polarization The science of fake news- Addressing fake news requires a multidisciplinary effort Societal mass polarisation is accelerating. In turn, this is wreaking havoc on our ability to find mutual agreement with those that hold a different viewpoint. Researchers conducted an experiment to explore this. A randomly assigned a sample of Twitter users were asked to follow a “bot” account that would share content from the opposing side of the political divide. Rather than decreasing polarization, they found that exposure to opposing viewpoints pushed people even further toward their own side. Similarly, individuals tend not to question the credibility of information when it aligned with our beliefs with the values. 4. How Politics limits Environmental Activism Trending Papers: Political power and renewable energy futures A critical review Environmental conflicts and defenders — A global overview These last two papers discuss a single idea: communities should be able to govern how private organisations and government use their environments. Both ideas make the argument for the means to control their environments to improve their freedoms. For example, if a group can govern how its energy is produced (locally rather than centrally/renewable vs non-renewable), this strengthens a community through the creation of new infrastructure and local economic opportunity. It delivers what they call an ‘Energy Democracy’, and allows for new types of negotiation power with speaking with a central body like government. BONUS: How did I find these papers? Typically to find out how influential something is within the academic community, you would use a metric like the citation number. Although citations are a reasonably good indicator of how impactful a paper might be, they can take anywhere from months to years to appear. It’s for this reason I am using the Mendeley Reader number as an indicator. It indicates how many academics have added a specific paper to their library. Full disclosure : I work as a product manager at Mendeley, so I am biased towards using the Mendeley metric. In theory, something like the number of tweets or blog posts for a given paper could also work. I recorded this number every month and was able to filter it down by academic discipline. After this, I plotted the reader data on a time series and calculated a score for the linear gradient and exponent. This score would tell me if readership was increasing or decreasing over time, aka — What’s trending? If you think you would find access to this data interesting, please let me know. Also published here .

What Sociologists Are Saying About 2020's Myriad Political Movements

SINGAPORE, Singapore, July 23rd, 2024/Chainwire/--A groundbreaking moment in decentralized finance (DeFi) is here as Magnify Cash unveils its state-of-the-art DeFi protocol and announces the upcoming Fair Launch of its native token, $MAG. As a pioneer in financial technology, Magnify Cash is set to reshape the financial services landscape for individuals and enterprises alike. The Future of Decentralized Credit Markets Magnify Cash is spearheading the next evolution in decentralized credit markets, unlocking liquidity for Real-World Assets (RWAs), Future Yield Tokens (FYTs), and a myriad of digital assets through its innovative, permissionless lending desks. Solving Key Financial Challenges Limited Liquidity: Bridging the gap for diverse digital assets.
Decentralization: Eliminating centralized and custodial lending limitations. Innovative Products and Solutions Permissionless Lending Desks: Facilitate borrowing and lending for any ERC-20 token against RWAs, FYTs, and other digital assets.
Decentralized Borrowing Options: Provide lending solutions without intermediaries, ensuring maximum security and transparency. Universal Collateral and Currency Support Magnify Cash’s Universal Collateral Acceptance feature supports an extensive array of tokenized assets, including ERC-1155 and ERC-721 tokens. Users can now leverage NFTs and tokenized real-world assets such as art, real estate, bonds, and luxury goods as collateral, unlocking liquidity without the need to liquidate these valuable assets. Key Features Comprehensive Support: Universal support for lending currencies and loan collateral (ERC-20, ERC-1155, ERC-721).
Seamless Operations: Asset-backed borrowing and lending with permissionless creation and management of lending desks.
Enhanced Liquidity: Extensive liquidity options for all digital assets, including stablecoins, ETH, liquid staking tokens (LSTs), meme tokens, blue chip NFTs, and treasuries.
Advanced Risk Management: A dynamic interest rate model utilizing linear interpolation and scaling factors, eliminating the need for oracles. Robust Security and Compliance Guardian Protocol Audit: Ensuring top-notch security.
Gaslite Gas Optimization Audit: Maximizing operational efficiency. Proven Testnet Success Over 500 lending desks created.
More than 500 loans issued.
Nearly 5,000 unique users engaged. Strategic Alliances and Future Roadmap Magnify Cash collaborates with leading industry partners like LydianLabs, FjordFoundry, Jiraiya, Tailored, Blaize, 4am, and H3AT Studios, amplifying the platform's capabilities and outreach. Key security audits by Guardian Protocol and Gaslite reinforce Magnify Cash’s commitment to providing a secure and reliable platform. Currently Live on Base Testnet Exciting Upcoming Launches Base Mainnet: Coming Soon (Including RWAs)
Ethereum Mainnet: Coming Soon (Including RWAs)
Potential Layer 2 launches: Abstract Chain, Blast Chain, and more. Fair Launch of $MAG Token The $MAG token will be launched through Fjord Foundry’s liquidity bootstrapping protocol, ensuring a fair and transparent distribution. The token sale runs from July 23rd, 8:00 PM EST to July 27th, 8:00 AM EST, with over 70% of tokens allocated to the public. This equitable approach encourages broad participation and fosters a strong, engaged community. Active participants, including lenders, liquidity providers, and borrowers, are rewarded with $MAG tokens. Participation Opportunities Secure $MAG Tokens: The MAG token sale begins on July 23rd at 8:00 PM EST (July 24th, 8:00 AM Singapore Time). Participants are invited to join the Fair Launch at Fjord Foundry here. Explore the Ecosystem: Magnify Cash invites interested parties to discover its comprehensive ecosystem by visiting the Website. Access the Pitch Deck, engaging Intro Video, detailed Whitepaper, and technical Documentation. Engage with Our Community: Magnify Cash encourages joining its vibrant community for real-time updates and support. Follow us on Twitter, join the Telegram group, and participate in discussions on Discord. About Magnify Cash Magnify Cash is a non-custodial and permissionless decentralized credit market that empowers users to lend ERC-20 tokens and borrow against non-fungible tokens (NFTs) and Real-World Assets (RWAs) with unmatched flexibility and transparency. Committed to true decentralization, Magnify Cash ensures universal support for ERC-20 tokens as lending currencies and ERC-1155/ERC-721 assets as collateral. The protocol’s innovative approach to on-chain credit markets offers a fresh perspective in the evolving world of decentralized finance. For more information, users can visit https://Magnify.Cash or follow Magnify Cash on Twitter at X.com/MagnifyCash. Magnify Media Kit: Link. Contact Press Manager Magnify.Cash socials@magnify.cash This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here. SINGAPORE, Singapore, July 23rd, 2024/Chainwire/--A groundbreaking moment in decentralized finance (DeFi) is here as Magnify Cash unveils its state-of-the-art DeFi protocol and announces the upcoming Fair Launch of its native token, $MAG. As a pioneer in financial technology, Magnify Cash is set to reshape the financial services landscape for individuals and enterprises alike. The Future of Decentralized Credit Markets Magnify Cash is spearheading the next evolution in decentralized credit markets, unlocking liquidity for Real-World Assets (RWAs), Future Yield Tokens (FYTs), and a myriad of digital assets through its innovative, permissionless lending desks. Solving Key Financial Challenges Limited Liquidity: Bridging the gap for diverse digital assets. Decentralization: Eliminating centralized and custodial lending limitations. Limited Liquidity: Bridging the gap for diverse digital assets. Decentralization: Eliminating centralized and custodial lending limitations. Innovative Products and Solutions Permissionless Lending Desks: Facilitate borrowing and lending for any ERC-20 token against RWAs, FYTs, and other digital assets. Decentralized Borrowing Options: Provide lending solutions without intermediaries, ensuring maximum security and transparency. Permissionless Lending Desks: Facilitate borrowing and lending for any ERC-20 token against RWAs, FYTs, and other digital assets. Decentralized Borrowing Options: Provide lending solutions without intermediaries, ensuring maximum security and transparency. Universal Collateral and Currency Support Magnify Cash’s Universal Collateral Acceptance feature supports an extensive array of tokenized assets, including ERC-1155 and ERC-721 tokens. Users can now leverage NFTs and tokenized real-world assets such as art, real estate, bonds, and luxury goods as collateral, unlocking liquidity without the need to liquidate these valuable assets. Key Features Comprehensive Support: Universal support for lending currencies and loan collateral (ERC-20, ERC-1155, ERC-721). Seamless Operations: Asset-backed borrowing and lending with permissionless creation and management of lending desks. Enhanced Liquidity: Extensive liquidity options for all digital assets, including stablecoins, ETH, liquid staking tokens (LSTs), meme tokens, blue chip NFTs, and treasuries. Advanced Risk Management: A dynamic interest rate model utilizing linear interpolation and scaling factors, eliminating the need for oracles. Comprehensive Support: Universal support for lending currencies and loan collateral (ERC-20, ERC-1155, ERC-721). Seamless Operations: Asset-backed borrowing and lending with permissionless creation and management of lending desks. Enhanced Liquidity: Extensive liquidity options for all digital assets, including stablecoins, ETH, liquid staking tokens (LSTs), meme tokens, blue chip NFTs, and treasuries. Advanced Risk Management: A dynamic interest rate model utilizing linear interpolation and scaling factors, eliminating the need for oracles. Robust Security and Compliance Guardian Protocol Audit: Ensuring top-notch security. Gaslite Gas Optimization Audit: Maximizing operational efficiency. Guardian Protocol Audit : Ensuring top-notch security. Guardian Protocol Audit Guardian Protocol Audit Gaslite Gas Optimization Audit : Maximizing operational efficiency. Gaslite Gas Optimization Audit Gaslite Gas Optimization Audit Proven Testnet Success Over 500 lending desks created. More than 500 loans issued. Nearly 5,000 unique users engaged. Over 500 lending desks created. More than 500 loans issued. Nearly 5,000 unique users engaged. Strategic Alliances and Future Roadmap Magnify Cash collaborates with leading industry partners like LydianLabs , FjordFoundry , Jiraiya , Tailored , Blaize, 4am, and H3AT Studios , amplifying the platform's capabilities and outreach. Key security audits by Guardian Protocol and Gaslite reinforce Magnify Cash’s commitment to providing a secure and reliable platform. LydianLabs LydianLabs FjordFoundry FjordFoundry Jiraiya Jiraiya Tailored Tailored H3AT Studios H3AT Studios Currently Live on Base Testnet Base Testnet Exciting Upcoming Launches Base Mainnet: Coming Soon (Including RWAs) Ethereum Mainnet: Coming Soon (Including RWAs) Potential Layer 2 launches: Abstract Chain, Blast Chain, and more. Base Mainnet: Coming Soon (Including RWAs) Ethereum Mainnet: Coming Soon (Including RWAs) Potential Layer 2 launches: Abstract Chain, Blast Chain, and more. Fair Launch of $MAG Token The $MAG token will be launched through Fjord Foundry’s liquidity bootstrapping protocol, ensuring a fair and transparent distribution. The token sale runs from July 23rd, 8:00 PM EST to July 27th, 8:00 AM EST, with over 70% of tokens allocated to the public. This equitable approach encourages broad participation and fosters a strong, engaged community. Active participants, including lenders, liquidity providers, and borrowers, are rewarded with $MAG tokens. Participation Opportunities Secure $MAG Tokens: The MAG token sale begins on July 23rd at 8:00 PM EST (July 24th, 8:00 AM Singapore Time). Participants are invited to join the Fair Launch at Fjord Foundry here . here here Explore the Ecosystem: Magnify Cash invites interested parties to discover its comprehensive ecosystem by visiting the Website . Access the Pitch Deck, engaging Intro Video, detailed Whitepaper, and technical Documentation. Website Website Engage with Our Community: Magnify Cash encourages joining its vibrant community for real-time updates and support. Follow us on Twitter , join the Telegram group , and participate in discussions on Discord . Twitter Twitter Telegram group Telegram group Discord Discord About Magnify Cash Magnify Cash is a non-custodial and permissionless decentralized credit market that empowers users to lend ERC-20 tokens and borrow against non-fungible tokens (NFTs) and Real-World Assets (RWAs) with unmatched flexibility and transparency. Committed to true decentralization, Magnify Cash ensures universal support for ERC-20 tokens as lending currencies and ERC-1155/ERC-721 assets as collateral. Magnify Cash Magnify Cash The protocol’s innovative approach to on-chain credit markets offers a fresh perspective in the evolving world of decentralized finance. For more information, users can visit https://Magnify.Cash or follow Magnify Cash on Twitter at X.com/MagnifyCash . https://Magnify.Cash https://Magnify.Cash X.com/MagnifyCash X.com/MagnifyCash Magnify Media Kit: Link . Link Link Contact Press Manager Magnify.Cash socials@magnify.cash This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here. This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here . here here

Magnify Cash Launches DeFi Protocol And Announces $MAG Token Fair Launch

ম্যাগনিফাই ক্যাশ ডিফাই প্রোটোকল চালু করেছে এবং $MAG টোকেন ফেয়ার লঞ্চের ঘোষণা করেছে

Magnify Cash führt DeFi-Protokoll ein und kündigt fairen Start des $MAG-Tokens an

Magnify Cash lanza el protocolo DeFi y anuncia el lanzamiento de la feria de tokens $MAG

Magnify Cash lance le protocole DeFi et annonce le lancement du salon des jetons $MAG

मैग्नीफाई कैश ने DeFi प्रोटोकॉल लॉन्च किया और $MAG टोकन फेयर लॉन्च की घोषणा की

Magnify CashがDeFiプロトコルを立ち上げ、$MAGトークンフェアのローンチを発表

Magnify Cash, DeFi 프로토콜 출시 및 $MAG 토큰 박람회 출시 발표

Magnify Cash lança protocolo DeFi e anuncia lançamento da $MAG Token Fair

Magnify Cash запускает протокол DeFi и объявляет о честном запуске токена $MAG

Magnify Cash, DeFi Protokolünü Başlatıyor ve $MAG Token Fuarı Lansmanını Duyurdu

Magnify Cash ra mắt giao thức DeFi và công bố ra mắt công bằng token $MAG

Magnify Cash 推出 DeFi 协议并宣布 $MAG 代币公平发行

If something is valuable, you need to protect it with trustworthy shields. This holds true for both tangible and digital assets. With hackers and malicious software constantly weaving their way through the web, cybersecurity specialists are on a mission to develop next-gen hazard mitigation methods to keep digital threats at bay. Durga Prasada Rao Sanagana, the lead cybersecurity architect at a renowned financial organization, is among the tech experts stepping up to combat cyber threats with advanced firewall defenses. “As technology for good evolves, so does ill-intentioned technology. Threats evolve faster than ever, so our defenses must keep pace, anticipate the next move, and outsmart the adversary,” says Sanagana, reflecting on the importance of revamping cybersecurity ahead of the curve. The North Star of Cybersecurity Innovation Thanks to his significant career strides, Sanagana has become a notable figure in cybersecurity. As a lead security architect for a major financial services company, Sanagana focuses on safeguarding digital assets and developing new defense methods. Applying his forte to real-life situations, Sanagana has built specialization in a myriad of sectors. This includes cybersecurity architecture and engineering, AI, data science, cloud security & architecture, and DevSecOps automation among others. His significant contributions in retail, banking and finance, pharmaceuticals, and healthcare are evident in the solutions he developed for global giants. His work today directly sparks a movement with the next generation of cybersecurity champions. When he is not actively conceptualizing new fundamentals for online safety, he mentors an international faculty development program and guides 140 faculty members, PhD scholars, and industry professionals in building a resilient cybersecurity architecture for data science operations. Passing the baton of knowledge certifies the continued success of the cybersecurity field. Beyond his direct career-related achievements, Sanagana participates actively in the cybersecurity community. That said, he has taken on an off-duty role as a peer reviewer for Springer and IEEE (Institute of Electrical and Electronics Engineers) conferences and regularly judges hackathons and international award events. Sanagana has also put his prowess to paper through notable pieces in Elsevier and JSTAR. Before and After: Perpetual Cybersecurity Growth In 2023, the cost of cybercrime in the United States saw an immense uptick. The increasing frequency and sophistication of cyber threats drove this significant growth. Sanagana's role is crucial in cultivating this aspect of the digital world. His projects surrounding next-gen firewalls, threat intelligence, threat modeling, and intrusion detection systems are fueled by deep learning techniques to predict and mitigate cyber threats. These initiatives offer a significant edge over traditional methods. Moreover, AI is applicable in nearly every industry, and cybersecurity is no exception. Sanagana recognizes the power of AI and stands at the forefront of integrating it into cybersecurity. The architect’s AI-driven adaptive learning systems for business intelligence simulation structures use machine learning and data analytics to enhance threat detection and proactive defense mechanisms. Locked and Loaded: 21st Century Cybersecurity The Internet of Things (IoT) is expanding by the minute, giving cyber threats a more spacious platform to attack. Sanagana envisions a multi-layered cybersecurity method combining AI, blockchain, and quantum computing in a solid attempt to thwart this. As cyber threats grow in complexity, sophisticated defenses are indispensable. Sanagana's work personifies the forward-thinking perspectives required to secure humanity's digital future. “The cyber threats we face in the present are in a constant state of change. The forms they take today may be completely unrecognizable by tomorrow. Our response must be equally innovative, integrating the best of technology and human ingenuity,” he concludes, highlighting his commitment to a safer digital landscape. Building next-gen firewalls that keep assets safe and hackers out; this is the focal point of Durga Prasada Rao Sanagana's work. This piece was published as part of HackerNoon’s Business Blogging Program. For more details, please click here. If something is valuable, you need to protect it with trustworthy shields. This holds true for both tangible and digital assets. With hackers and malicious software constantly weaving their way through the web, cybersecurity specialists are on a mission to develop next-gen hazard mitigation methods to keep digital threats at bay. If something is valuable, you need to protect it with trustworthy shields. This holds true for both tangible and digital assets. With hackers and malicious software constantly weaving their way through the web, cybersecurity specialists are on a mission to develop next-gen hazard mitigation methods to keep digital threats at bay. Durga Prasada Rao Sanagana , the lead cybersecurity architect at a renowned financial organization, is among the tech experts stepping up to combat cyber threats with advanced firewall defenses. Durga Prasada Rao Sanagana Durga Prasada Rao Sanagana “As technology for good evolves, so does ill-intentioned technology. Threats evolve faster than ever, so our defenses must keep pace, anticipate the next move, and outsmart the adversary,” says Sanagana, reflecting on the importance of revamping cybersecurity ahead of the curve. “As technology for good evolves, so does ill-intentioned technology. Threats evolve faster than ever, so our defenses must keep pace, anticipate the next move, and outsmart the adversary,” The North Star of Cybersecurity Innovation The North Star of Cybersecurity Innovation Thanks to his significant career strides, Sanagana has become a notable figure in cybersecurity. As a lead security architect for a major financial services company, Sanagana focuses on safeguarding digital assets and developing new defense methods. Applying his forte to real-life situations, Sanagana has built specialization in a myriad of sectors. This includes cybersecurity architecture and engineering, AI, data science, cloud security & architecture, and DevSecOps automation among others. His significant contributions in retail, banking and finance, pharmaceuticals, and healthcare are evident in the solutions he developed for global giants. His work today directly sparks a movement with the next generation of cybersecurity champions. When he is not actively conceptualizing new fundamentals for online safety, he mentors an international faculty development program and guides 140 faculty members, PhD scholars, and industry professionals in building a resilient cybersecurity architecture for data science operations. Passing the baton of knowledge certifies the continued success of the cybersecurity field. Beyond his direct career-related achievements, Sanagana participates actively in the cybersecurity community. That said, he has taken on an off-duty role as a peer reviewer for Springer and IEEE (Institute of Electrical and Electronics Engineers) conferences and regularly judges hackathons and international award events. Sanagana has also put his prowess to paper through notable pieces in Elsevier and JSTAR. Before and After: Perpetual Cybersecurity Growth Before and After: Perpetual Cybersecurity Growth In 2023, the cost of cybercrime in the United States saw an immense uptick. The increasing frequency and sophistication of cyber threats drove this significant growth. Sanagana's role is crucial in cultivating this aspect of the digital world. His projects surrounding next-gen firewalls, threat intelligence, threat modeling, and intrusion detection systems are fueled by deep learning techniques to predict and mitigate cyber threats. These initiatives offer a significant edge over traditional methods. Moreover, AI is applicable in nearly every industry, and cybersecurity is no exception. Sanagana recognizes the power of AI and stands at the forefront of integrating it into cybersecurity. The architect’s AI-driven adaptive learning systems for business intelligence simulation structures use machine learning and data analytics to enhance threat detection and proactive defense mechanisms. Locked and Loaded: 21st Century Cybersecurity Locked and Loaded: 21st Century Cybersecurity The Internet of Things (IoT) is expanding by the minute, giving cyber threats a more spacious platform to attack. Sanagana envisions a multi-layered cybersecurity method combining AI, blockchain, and quantum computing in a solid attempt to thwart this. As cyber threats grow in complexity, sophisticated defenses are indispensable. Sanagana's work personifies the forward-thinking perspectives required to secure humanity's digital future. “The cyber threats we face in the present are in a constant state of change. The forms they take today may be completely unrecognizable by tomorrow. Our response must be equally innovative, integrating the best of technology and human ingenuity,” he concludes, highlighting his commitment to a safer digital landscape. “The cyber threats we face in the present are in a constant state of change. The forms they take today may be completely unrecognizable by tomorrow. Our response must be equally innovative, integrating the best of technology and human ingenuity,” Building next-gen firewalls that keep assets safe and hackers out; this is the focal point of Durga Prasada Rao Sanagana's work. This piece was published as part of HackerNoon’s Business Blogging Program. For more details, please click here. This piece was published as part of HackerNoon’s Business Blogging Program. For more details, please click here . here

Tech Expert Durga Sanagana Advances Next-Gen Firewalls and Threat Modeling Techniques

টেক এক্সপার্ট দুর্গা সনাগানা নেক্সট-জেন ফায়ারওয়াল এবং থ্রেট মডেলিং কৌশলকে অগ্রসর করেছে

Tech-Expertin Durga Sanagana entwickelt Firewalls und Bedrohungsmodellierungstechniken der nächsten Generation weiter

El experto en tecnología Durga Sanagana avanza en las técnicas de modelado de amenazas y firewalls de próxima generación

L'expert technique Durga Sanagana fait progresser les pare-feu de nouvelle génération et les techniques de modélisation des menaces

तकनीकी विशेषज्ञ दुर्गा सनागना ने अगली पीढ़ी की फायरवॉल और खतरा मॉडलिंग तकनीक विकसित की

技術専門家のドゥルガ・サナガナが次世代ファイアウォールと脅威モデリング技術を進化させる

기술 전문가 Durga Sanagana, 차세대 방화벽 및 위협 모델링 기술 발전

A especialista em tecnologia Durga Sanagana desenvolve firewalls de última geração e técnicas de modelagem de ameaças

Технический эксперт Дурга Санагана продвигает межсетевые экраны нового поколения и методы моделирования угроз

Teknoloji Uzmanı Durga Sanagana Yeni Nesil Güvenlik Duvarlarını ve Tehdit Modelleme Tekniklerini Geliştiriyor

Chuyên gia công nghệ Durga Sanagana nâng cao kỹ thuật mô hình hóa mối đe dọa và tường lửa thế hệ tiếp theo

技术专家 Durga Sanagana 推进下一代防火墙和威胁建模技术

The age of the Internet began in earnest in April 1993 when it was released into the public domain. Since then, it has grown to become an all-pervading force in our lives, impacting the way we communicate, carry out business and banking transactions, entertain ourselves, spend our leisure time, and even order our food and groceries. All of this was inconceivable not all that long ago, as would be well remembered by the Boomers, Generation X, and even the early Millennials. Though many of them view the pre-digital era that existed till the very early Nineties with rose-tinted nostalgia, one can generally agree that the World Wide Web has done far more good than bad. That the Internet changed life as mankind knew it for thousands of years in a fundamental manner is a fact. It ushered in the digital revolution, which is possibly even more path-breaking in its impact on mankind than the harnessing of fire and the invention of the wheel. We are, in fact, in the throes of a digital revolution that has only about begun. If things weren't interesting enough for us, even as we see our lives transform at a breakneck speed thanks to the changes ushered in by the entry of the Internet into our lives, we need to brace ourselves for another epochal transformation caused by the advent of AI or Artificial Intelligence technology. Where AI is different from the technologies preceding it is in the fact that it is already autonomous of human oversight to some extent in its decision-making and may one day become fully independent, which has led to alarm among many, who feat that its large-scale adoption might lead to a dystopian nightmare straight out of an HG Wells sci-fi thriller. People fear that AI could render everybody from writers and surgeons to lawyers and teachers redundant, leading to an existential crisis on an epic scale. For all the fear-mongering around AI, which admittedly appears justified even to the brightest of human minds, many of whom have expressed their concerns about its dangers, it does offer us myriad benefits across several very practical applications. From preventing fraud and analyzing risks in the investment sector to carrying out complex scientific research by perusing complex data and from automating repetitive mundane tasks to transforming healthcare with its varied abilities, including interpreting data to help identify diseases and assisting in drug development, AI can be of immense service to mankind. AI is still essentially at a very nascent stage and many years away from a time where it is unilaterally and independently able to make major life-and-death decisions of its own volition. That is not to say that it never may, which is why governments, lawmakers, technologists, scientists and philosophers must utilise the time we have to regulate the technology and decide on the ethical framework within which it must be deployed. That the AI age has dawned upon us is a fact, but whether it will turn out to be all pervasive and more is something that we don't yet have the answer to. The Dot Com bubble, for instance, had nearly sunk the Internet revolution until the proponents of the technology proved its usefulness in many other ways. Who knows when the current rash of AI startups pitching a gazillion AI products might go belly up, making people wary of the technology to the extent that they may want to dump it? Alternatively, it may see its ups and downs and finally become the new normal in a much better-regulated and rationalized manner. In any case, we are in for an exciting time ahead. The age of the Internet began in earnest in April 1993 when it was released into the public domain. Since then, it has grown to become an all-pervading force in our lives, impacting the way we communicate, carry out business and banking transactions, entertain ourselves, spend our leisure time, and even order our food and groceries. All of this was inconceivable not all that long ago, as would be well remembered by the Boomers, Generation X, and even the early Millennials. Though many of them view the pre-digital era that existed till the very early Nineties with rose-tinted nostalgia, one can generally agree that the World Wide Web has done far more good than bad. That the Internet changed life as mankind knew it for thousands of years in a fundamental manner is a fact. It ushered in the digital revolution, which is possibly even more path-breaking in its impact on mankind than the harnessing of fire and the invention of the wheel. We are, in fact, in the throes of a digital revolution that has only about begun. If things weren't interesting enough for us, even as we see our lives transform at a breakneck speed thanks to the changes ushered in by the entry of the Internet into our lives, we need to brace ourselves for another epochal transformation caused by the advent of AI or Artificial Intelligence technology. Where AI is different from the technologies preceding it is in the fact that it is already autonomous of human oversight to some extent in its decision-making and may one day become fully independent, which has led to alarm among many, who feat that its large-scale adoption might lead to a dystopian nightmare straight out of an HG Wells sci-fi thriller. People fear that AI could render everybody from writers and surgeons to lawyers and teachers redundant, leading to an existential crisis on an epic scale. For all the fear-mongering around AI, which admittedly appears justified even to the brightest of human minds, many of whom have expressed their concerns about its dangers, it does offer us myriad benefits across several very practical applications. From preventing fraud and analyzing risks in the investment sector to carrying out complex scientific research by perusing complex data and from automating repetitive mundane tasks to transforming healthcare with its varied abilities, including interpreting data to help identify diseases and assisting in drug development, AI can be of immense service to mankind. AI is still essentially at a very nascent stage and many years away from a time where it is unilaterally and independently able to make major life-and-death decisions of its own volition. That is not to say that it never may, which is why governments, lawmakers, technologists, scientists and philosophers must utilise the time we have to regulate the technology and decide on the ethical framework within which it must be deployed. That the AI age has dawned upon us is a fact, but whether it will turn out to be all pervasive and more is something that we don't yet have the answer to. The Dot Com bubble, for instance, had nearly sunk the Internet revolution until the proponents of the technology proved its usefulness in many other ways. Who knows when the current rash of AI startups pitching a gazillion AI products might go belly up, making people wary of the technology to the extent that they may want to dump it? Alternatively, it may see its ups and downs and finally become the new normal in a much better-regulated and rationalized manner. In any case, we are in for an exciting time ahead.

From the Age of the Internet to the Age of AI

I used to believe that software for creators was somewhat oversaturated — and upon doing more research, I found out that I was wrong. The Evolution of the Creator Economy First, let’s get the definition out of the way — the “Creator Economy” is not about dancing TikTokers or Instagram influencers partying on yachts. In fact, the largest share of content creators are small businesses. Thus, the “Creator Economy” refers to independent businesses and side-hustles launched by individuals who are looking to earn an income off of their knowledge and skills through digital mediums. To give you an idea of how fast the sector has grown since 2011, take a look at how much creators have earned on Gumroad as of 2020 👇 This growth is being driven by: how we create and consume content
access to new software and media platforms that empower creators
new ways of making money from your work Social Media and The Rise of Creators Initially, social media enabled people to build large audiences and monetization came from brand deals or view count. In this old web economy, earning a living from your work was only possible if you could reach millions. Therefore, it was all dependent on scale. The Move towards Direct Monetization Through social media, creators owned nothing, their content belonged to the platforms. As a result, creators began splitting their efforts across multiple mediums (e.g.newsletters, courses, podcasts) to monetize their audiences directly, without being solely reliant on brand deals and ad revenue. This is when things started shifting. Creators as Owners Through alternative ways to monetize their audience, creators began to realize if they owned the distribution, they could monetize much more differently. In the old world (e.g. instagram, youtube, tiktok), the economics were different. You previously needed thousands, if not millions, of followers to earn $1,000/month. Through direct monetization on new platforms (e.g. Twitch, Discord, OnlyFans, Patreon) you could achieve the same result with a much smaller audience. More and more of these new platforms are offering new ways for creators to make money by going straight to the end-user and offering something of value in exchange for their money. The Creator Economy Landscape Goldman Sachs estimates that the creator economy will double to $480 billion by 2027. Technology has and will continue to play a huge role in this, as it has lowered the barriers to content creation. Individuals are now free to choose which kind of work they take on (e.g. newsletters, livestreams, podcasts). “Earning a living as a creator is an evolution of the so-called gig economy facilitated by companies like Uber and TaskRabbit.”
-Sam Yam, Co-Founder of Patreon This has fueled demand, not only for content platforms, but also new services and tools that help creators manage their business. The Opportunity for Startups in the Creator Economy The shift towards diversified revenue streams and content ownership has resulted in the birth of a myriad of solutions and opportunities for creators, founders, and VCs. Creators have access to new technologies and platforms through which they can engage their audiences and monetize their content. Founders and VCs, recognizing the value of direct audience relationships, have focused on investing and developing tools to facilitate content creation, distribution, and monetization. Some of the business models that VCs are paying attention to are: B2B Services: this includes analytics tools, tools for managing communities and fan relationships, and services that help creators with operations (e.g. legal, accounting, or financing)


Premium Fan Experiences: platforms that facilitate the creation of unique, high-touch experiences for superfans. These can include virtual meet-and-greets, personal shout-outs, exclusive content, and more.


Education and Courses: Creators with specific expertise in can develop and sell courses or offer coaching and consulting services. One of the most overlooked characteristics of the creator economy is that creators are in fact media companies, and these companies require services, tech, and infrastructure to operate. So, do I think startups should build more software for creators? Definitely, yes. If you enjoyed this article, and you’d like to see more content like this, sign up for my newsletter here! I used to believe that software for creators was somewhat oversaturated — and upon doing more research, I found out that I was wrong. The Evolution of the Creator Economy First, let’s get the definition out of the way — the “Creator Economy” is not about dancing TikTokers or Instagram influencers partying on yachts. In fact, the largest share of content creators are small businesses. Thus, the “Creator Economy” refers to independent businesses and side-hustles launched by individuals who are looking to earn an income off of their knowledge and skills through digital mediums. To give you an idea of how fast the sector has grown since 2011, take a look at how much creators have earned on Gumroad as of 2020 👇 This growth is being driven by: how we create and consume content access to new software and media platforms that empower creators new ways of making money from your work how we create and consume content access to new software and media platforms that empower creators new ways of making money from your work Social Media and The Rise of Creators Initially, social media enabled people to build large audiences and monetization came from brand deals or view count. In this old web economy, earning a living from your work was only possible if you could reach millions. Therefore, it was all dependent on scale. The Move towards Direct Monetization Through social media, creators owned nothing, their content belonged to the platforms. As a result, creators began splitting their efforts across multiple mediums (e.g.newsletters, courses, podcasts) to monetize their audiences directly, without being solely reliant on brand deals and ad revenue. This is when things started shifting. Creators as Owners Through alternative ways to monetize their audience, creators began to realize if they owned the distribution, they could monetize much more differently. In the old world (e.g. instagram, youtube, tiktok), the economics were different. You previously needed thousands, if not millions, of followers to earn $1,000/month. Through direct monetization on new platforms (e.g. Twitch, Discord, OnlyFans, Patreon) you could achieve the same result with a much smaller audience. More and more of these new platforms are offering new ways for creators to make money by going straight to the end-user and offering something of value in exchange for their money. The Creator Economy Landscape Goldman Sachs estimates that the creator economy will double to $480 billion by 2027. Goldman Sachs Goldman Sachs Technology has and will continue to play a huge role in this, as it has lowered the barriers to content creation. Individuals are now free to choose which kind of work they take on (e.g. newsletters, livestreams, podcasts). “Earning a living as a creator is an evolution of the so-called gig economy facilitated by companies like Uber and TaskRabbit.” -Sam Yam, Co-Founder of Patreon “Earning a living as a creator is an evolution of the so-called gig economy facilitated by companies like Uber and TaskRabbit.” “Earning a living as a creator is an evolution of the so-called gig economy facilitated by companies like Uber and TaskRabbit.” -Sam Yam, Co-Founder of Patreon This has fueled demand, not only for content platforms, but also new services and tools that help creators manage their business. The Opportunity for Startups in the Creator Economy The Opportunity for Startups in the Creator Economy The shift towards diversified revenue streams and content ownership has resulted in the birth of a myriad of solutions and opportunities for creators, founders, and VCs. Creators have access to new technologies and platforms through which they can engage their audiences and monetize their content. Founders and VCs, recognizing the value of direct audience relationships, have focused on investing and developing tools to facilitate content creation, distribution, and monetization. Some of the business models that VCs are paying attention to are: paying attention to paying attention to B2B Services: this includes analytics tools, tools for managing communities and fan relationships, and services that help creators with operations (e.g. legal, accounting, or financing) Premium Fan Experiences: platforms that facilitate the creation of unique, high-touch experiences for superfans. These can include virtual meet-and-greets, personal shout-outs, exclusive content, and more. Education and Courses: Creators with specific expertise in can develop and sell courses or offer coaching and consulting services. B2B Services: this includes analytics tools, tools for managing communities and fan relationships, and services that help creators with operations (e.g. legal, accounting, or financing) B2B Services: this includes a nalytics tools, tools for managing communities and fan relationships, and services that help creators with operations (e.g. legal, accounting, or financing) B2B Services: a Premium Fan Experiences: platforms that facilitate the creation of unique, high-touch experiences for superfans. These can include virtual meet-and-greets, personal shout-outs, exclusive content, and more. Premium Fan Experiences: platforms that facilitate the creation of unique, high-touch experiences for superfans. These can include virtual meet-and-greets, personal shout-outs, exclusive content, and more. Premium Fan Experiences: Education and Courses: Creators with specific expertise in can develop and sell courses or offer coaching and consulting services. Education and Courses: Creators with specific expertise in can develop and sell courses or offer coaching and consulting services. Education and Courses: One of the most overlooked characteristics of the creator economy is that creators are in fact media companies, and these companies require services, tech, and infrastructure to operate. So, do I think startups should build more software for creators? Definitely, yes. If you enjoyed this article, and you’d like to see more content like this, sign up for my newsletter here ! here here

Are Startups Building Enough for the Creator Economy?

The GDPR framework was introduced in Europe in May 2018 and has since become a global reference for protecting user data and privacy in today’s digital society… In the generative AI era, ChatpGPT, Claude 3, Gemini, and Perplexity AI are major players. Data privacy is crucial, and we need to move from the hype stage to the generative AI privacy focus. While digital tools are valuable, privacy is even more essential in these turbulent times, as hackers continue to discover new ways to exploit vulnerabilities. 1. What Is Generative AI? Generative AI refers to artificial intelligence capable of creating new content, including words, images, music, code, video, art… In this exciting time, people with fewer skills in writing can learn how to write, and those with fewer skills in drawing can generate amazing art. In coding, junior developers can develop their skills with a wealth of knowledge. 2. The GDPR and Data Redaction 2018 was the year the GDPR was enacted in Europe; data is precious it can be in motion, at rest, and in use. One of the primary requirements of the framework is the right to privacy and the secure processing of personal data. Data redaction is paramount in protecting sensitive information for better privacy. Currently, some services allow developers to redact data in every project they are getting involved in. The advantages are**:** Applications can be shipped with privacy in mind. Users will adopt new tools in the generative AI with trust over the hype. Insecurity costs more than security. Once data is redacted with powerful tools, it becomes difficult for black hat hackers to discover sensitive data easily. With the right tools, we can avoid data theft and breaches for better privacy. 3. Why Data Redaction With Generative AI Tools? Generative AI tools have access to a vast amount of knowledge derived from millions to billions of pieces of trained data. However, if security measures are not properly implemented with a secure-by-design approach, these tools can be vulnerable to exploitation. Once integrated into apps, they can potentially be used for data theft. On the other hand, powerful AI tools can also be used to redact sensitive information. Additionally, companies are entering the market to address security issues by offering a set of APIs. With a set of security APIs, we can redact data for better user privacy, and there are advantages to enumerate such as: Efficiency: With automated data redaction, we can reduce the time for manual tasks by gaining more time. Precision: with a myriad of AI-driven tools, they can detect and redact sensitive information. Scalability: Generative AI tools can handle a large amount of data, and once integrated into apps, they can receive thousands of APIs requests daily; it’s an opportunity for several organizations to process large amounts of data. 4. How to Redact Data Using GDPR-DRIVEN Services Data redaction is crucial in today’s society; people desire enhanced privacy, not just a slogan! Companies that prioritize customers invest their efforts in data protection. Here are the steps for data redaction: 1. Identify and Classify Sensitive Data Machine learning plays a valuable role in identifying and classifying sensitive data within large datasets. Technology is incomplete without data! The strength of algorithms lies in their ability to be trained to detect personally identifiable information, such as phone numbers, street addresses, IP addresses, and financial information. 2. Generative AI Model Implementation By automating the redaction process, the models can replace sensitive data with non-sensitive information; it can be used as an alternative when needed. 3. Validate Outputs GDPR is the primary framework for regulating data privacy. The process involves validating redacted data for accuracy to prevent leakage of sensitive information. Continuous monitoring is required. 4. Use GDPR-DRIVEN Services Data privacy should not just be a slogan. With the numerous services available today from various vendors, it's possible to integrate a set of APIs into every project with just a few lines of code. These services can include audit logs, data redaction, encryption, and access control to enhance security. Conclusion Data privacy should not be just a slogan. With the multitude of services available, we can redact data to enhance privacy. In the era of generative AI, there will likely be more attacks from black hat hackers who actively learn and understand user behavior using AI tools. They can exploit vulnerabilities and harm users by using malware, rootkits, and keyloggers to steal data. Using GDPR-compliant services is a great way to ensure that data is redacted and that users are protected from harmful attacks. The GDPR framework was introduced in Europe in May 2018 and has since become a global reference for protecting user data and privacy in today’s digital society… GDPR GDPR In the generative AI era, ChatpGPT, Claude 3, Gemini, and Perplexity AI are major players. Data privacy is crucial, and we need to move from the hype stage to the generative AI privacy focus. While digital tools are valuable, privacy is even more essential in these turbulent times, as hackers continue to discover new ways to exploit vulnerabilities. 1. What Is Generative AI? Generative AI refers to artificial intelligence capable of creating new content, including words, images, music, code, video, art… In this exciting time, people with fewer skills in writing can learn how to write, and those with fewer skills in drawing can generate amazing art. In coding, junior developers can develop their skills with a wealth of knowledge. 2. The GDPR and Data Redaction 2018 was the year the GDPR was enacted in Europe; data is precious it can be in motion, at rest, and in use. One of the primary requirements of the framework is the right to privacy and the secure processing of personal data. Data redaction is paramount in protecting sensitive information for better privacy. Currently, some services allow developers to redact data in every project they are getting involved in. The advantages are**:** Applications can be shipped with privacy in mind. Applications can be shipped with privacy in mind. Users will adopt new tools in the generative AI with trust over the hype. Users will adopt new tools in the generative AI with trust over the hype. Insecurity costs more than security. Once data is redacted with powerful tools, it becomes difficult for black hat hackers to discover sensitive data easily. Insecurity costs more than security. Once data is redacted with powerful tools, it becomes difficult for black hat hackers to discover sensitive data easily. With the right tools, we can avoid data theft and breaches for better privacy. With the right tools, we can avoid data theft and breaches for better privacy. 3. Why Data Redaction With Generative AI Tools? Generative AI tools have access to a vast amount of knowledge derived from millions to billions of pieces of trained data. However, if security measures are not properly implemented with a secure-by-design approach, these tools can be vulnerable to exploitation. Once integrated into apps, they can potentially be used for data theft. On the other hand, powerful AI tools can also be used to redact sensitive information. Additionally, companies are entering the market to address security issues by offering a set of APIs. With a set of security APIs, we can redact data for better user privacy, and there are advantages to enumerate such as: Efficiency: With automated data redaction, we can reduce the time for manual tasks by gaining more time. Efficiency : With automated data redaction, we can reduce the time for manual tasks by gaining more time. Efficiency Precision: with a myriad of AI-driven tools, they can detect and redact sensitive information. Precision : with a myriad of AI-driven tools, they can detect and redact sensitive information. Precision Scalability: Generative AI tools can handle a large amount of data, and once integrated into apps, they can receive thousands of APIs requests daily; it’s an opportunity for several organizations to process large amounts of data. Scalability : Generative AI tools can handle a large amount of data, and once integrated into apps, they can receive thousands of APIs requests daily; it’s an opportunity for several organizations to process large amounts of data. Scalability 4. How to Redact Data Using GDPR-DRIVEN Services Data redaction is crucial in today’s society; people desire enhanced privacy, not just a slogan! Companies that prioritize customers invest their efforts in data protection. Here are the steps for data redaction: 1. Identify and Classify Sensitive Data Identify and Classify Sensitive Data Machine learning plays a valuable role in identifying and classifying sensitive data within large datasets. Technology is incomplete without data! The strength of algorithms lies in their ability to be trained to detect personally identifiable information, such as phone numbers, street addresses, IP addresses, and financial information. 2. Generative AI Model Implementation By automating the redaction process, the models can replace sensitive data with non-sensitive information; it can be used as an alternative when needed. 3. Validate Outputs GDPR is the primary framework for regulating data privacy. The process involves validating redacted data for accuracy to prevent leakage of sensitive information. Continuous monitoring is required. 4. Use GDPR-DRIVEN Services Data privacy should not just be a slogan. With the numerous services available today from various vendors, it's possible to integrate a set of APIs into every project with just a few lines of code. These services can include audit logs, data redaction, encryption, and access control to enhance security. Conclusion Data privacy should not be just a slogan. With the multitude of services available, we can redact data to enhance privacy. In the era of generative AI, there will likely be more attacks from black hat hackers who actively learn and understand user behavior using AI tools. They can exploit vulnerabilities and harm users by using malware, rootkits, and keyloggers to steal data. Using GDPR-compliant services is a great way to ensure that data is redacted and that users are protected from harmful attacks.

Generative AI: How to Redact Data With GDPR Driven Services

Authors:
(1) Ronen Eldan, Microsoft Research (email: roneneldan@microsoft.com);
(2) Mark Russinovich, Microsoft Azure and Both authors contributed equally to this work, (email: mark.russinovich@microsoft.com). Table of Links Abstract and Introduction
Description of our technique
Evaluation methodology
Results
Conclusion, Acknowledgment, and References
Appendix Abstract Large language models (LLMs) are trained on massive internet corpora that often contain copyrighted content. This poses legal and ethical challenges for the developers and users of these models, as well as the original authors and publishers. In this paper, we propose a novel technique for unlearning a subset of the training data from a LLM, without having to retrain it from scratch. We evaluate our technique on the task of unlearning the Harry Potter books from the Llama2-7b model (a generative language model recently open-sourced by Meta). While the model took over 184K GPU-hours to pretrain, we show that in about 1 GPU hour of finetuning, we effectively erase the model’s ability to generate or recall Harry Potter-related content, while its performance on common benchmarks (such as Winogrande, Hellaswag, arc, boolq and piqa) remains almost unaffected. To the best of our knowledge, this is the first paper to present an effective technique for unlearning in generative language models. Our technique consists of three main components: First, we use a reinforced model that is further trained on the target data to identify the tokens that are most related to the unlearning target, by comparing its logits with those of a baseline model. Second, we replace idiosyncratic expressions in the target data with generic counterparts, and leverage the model’s own predictions to generate alternative labels for every token. These labels aim to approximate the next-token predictions of a model that has not been trained on the target data. Third, we finetune the model on these alternative labels, which effectively erases the original text from the model’s memory whenever it is prompted with its context. 1 Introduction In the rapidly evolving domain of artificial intelligence and machine learning, Large Language Models (LLMs) stand as a testament to both our accomplishments and the challenges that lie ahead. Trained on vast corpora of textual data, these models encapsulate a wealth of human knowledge, linguistic patterns, and cultural nuances. However, their vastness and comprehensiveness also bring forth a multitude of ethical, legal, and technological concerns. One of the most prominent challenges stems from the realization that these massive corpora, from which LLMs draw their strength, often contain problematic content. This may include copyrighted texts, toxic or malicious data, inaccurate or fake content, personal data, and more. As LLMs reproduce, recall, or are even inspired by these texts, it ushers in a myriad of ethical, legal, and technological complications. Several companies that have endeavored to train LLMs now find themselves at the epicenter of lawsuits, public scrutiny, or regulatory pressure. Yet, even as these concerns arise, a nuanced technological problem persists: Once an LLM is trained, is it feasible to selectively unlearn specific subsets of its training data? Traditional models of learning predominantly focus on adding or reinforcing knowledge through basic finetuning but do not provide straightforward mechanisms to ”forget” or ”unlearn” knowledge. Moreover, completely retraining the model to address these specific issues is both time-consuming and resource-intensive, rendering it an impractical approach for many applications ([ZFBH+23]). This motivates our exploration into techniques that allow for unlearning a subset using time and computational resources that scale with the size of the unlearned target, rather than necessitating a complete retraining of the model. In this paper, we seek to address this challenge head-on. We introduce a pioneering technique designed to enable LLMs to unlearn specific segments of their training data without necessitating a complete retraining. Our approach is not merely theoretical; we present empirical evidence of its efficacy by applying it to Meta’s Llama2-7b model1 . As a proof of concept, we demonstrate that, while the original model can easily recover very detailed and nuanced information from the books, it’s possible for the model to essentially ”forget” the intricate narratives of the Harry Potter series ([Row07]), all while retaining its prowess on established benchmarks. To get a first impression of the fine-tuned model produced by our technique, Figure 1 compares the completions, on several prompts, of the baseline model (Llama2-7b-chat-hf) and a variant which has been fine-tuned for roughly 30 minutes on 4 A100-GPUs. Figure 2 compares the performance of these two models on some common benchmarks ([YBS19, CLC+19, ZHB+19, MCKS18, BHT+19, SLBBC19]) and Figure 3 compares the next token probability distributions for the sentence ”Harry Potter studies” over different steps of fine-tuning, showing how the most likely next token gradually shifts from ”magic” to generic completions. Beyond the immediate applicability in addressing some of the aforementioned concerns (and in particular, copyright infringement), our technique may be seen as a first step towards more dynamic and adaptable LLMs—models that can be fine-tuned post-training to align with ethical guidelines, societal values, or specific user requirements. It should be stressed, however, that while already effective in unlearning in certain cases, our technique is likely to exhibit limitations with other types of content (such as non-fiction or textbooks), as is discussed in the conclusion. Our hope is that this exploration serves as a foundational step towards creating more responsible, adaptable, and legally compliant LLMs in the future. 1.1 Related work While there is a growing body of work in the topic of unlearning in machine-learning in general (see [JLZ+22, NHN+22, ZNIS23] and references therein), the majority of works focus on classification tasks, while the literature concerning generative models or specifically LLMs is still quite slim. The very recent paper [ZFBH+23] highlights the related challenges and implications and discusses some high-level directions for potential mitigation. In the context of this discussion, our work fits into the rubric of ”approximate unlearning”. Recent works that propose concrete unlearning techniques for generative models are [JYY+22] which suggests a technique shown to address privacy risks in certain settings, and [WCY+23] which proposes an algorithm called knowledge-gap-alignment which may be in, certain cases, relevant for LLMs but relies on assumptions that do not seem to hold in our setting. This paper is available on arxiv under CC 4.0 license. 1 Our model can be found at https://huggingface.co/microsoft/Llama2-7b-WhoIsHarryPotter Authors: (1) Ronen Eldan, Microsoft Research (email: roneneldan@microsoft.com); (2) Mark Russinovich, Microsoft Azure and Both authors contributed equally to this work, (email: mark.russinovich@microsoft.com). Authors: Authors: (1) Ronen Eldan, Microsoft Research (email: roneneldan@microsoft.com); (2) Mark Russinovich, Microsoft Azure and Both authors contributed equally to this work, (email: mark.russinovich@microsoft.com). Table of Links Abstract and Introduction Description of our technique Evaluation methodology Results Conclusion, Acknowledgment, and References Appendix Abstract and Introduction Abstract and Introduction Description of our technique Description of our technique Evaluation methodology Evaluation methodology Results Results Conclusion, Acknowledgment, and References Conclusion, Acknowledgment, and References Appendix Appendix Abstract Large language models (LLMs) are trained on massive internet corpora that often contain copyrighted content. This poses legal and ethical challenges for the developers and users of these models, as well as the original authors and publishers. In this paper, we propose a novel technique for unlearning a subset of the training data from a LLM, without having to retrain it from scratch. We evaluate our technique on the task of unlearning the Harry Potter books from the Llama2-7b model (a generative language model recently open-sourced by Meta). While the model took over 184K GPU-hours to pretrain, we show that in about 1 GPU hour of finetuning, we effectively erase the model’s ability to generate or recall Harry Potter-related content, while its performance on common benchmarks (such as Winogrande, Hellaswag, arc, boolq and piqa) remains almost unaffected. To the best of our knowledge, this is the first paper to present an effective technique for unlearning in generative language models. Our technique consists of three main components: First, we use a reinforced model that is further trained on the target data to identify the tokens that are most related to the unlearning target, by comparing its logits with those of a baseline model. Second, we replace idiosyncratic expressions in the target data with generic counterparts, and leverage the model’s own predictions to generate alternative labels for every token. These labels aim to approximate the next-token predictions of a model that has not been trained on the target data. Third, we finetune the model on these alternative labels, which effectively erases the original text from the model’s memory whenever it is prompted with its context. 1 Introduction In the rapidly evolving domain of artificial intelligence and machine learning, Large Language Models (LLMs) stand as a testament to both our accomplishments and the challenges that lie ahead. Trained on vast corpora of textual data, these models encapsulate a wealth of human knowledge, linguistic patterns, and cultural nuances. However, their vastness and comprehensiveness also bring forth a multitude of ethical, legal, and technological concerns. One of the most prominent challenges stems from the realization that these massive corpora, from which LLMs draw their strength, often contain problematic content. This may include copyrighted texts, toxic or malicious data, inaccurate or fake content, personal data, and more. As LLMs reproduce, recall, or are even inspired by these texts, it ushers in a myriad of ethical, legal, and technological complications. Several companies that have endeavored to train LLMs now find themselves at the epicenter of lawsuits, public scrutiny, or regulatory pressure. Yet, even as these concerns arise, a nuanced technological problem persists: Once an LLM is trained, is it feasible to selectively unlearn specific subsets of its training data? Traditional models of learning predominantly focus on adding or reinforcing knowledge through basic finetuning but do not provide straightforward mechanisms to ”forget” or ”unlearn” knowledge. Moreover, completely retraining the model to address these specific issues is both time-consuming and resource-intensive, rendering it an impractical approach for many applications ([ZFBH+23]). This motivates our exploration into techniques that allow for unlearning a subset using time and computational resources that scale with the size of the unlearned target, rather than necessitating a complete retraining of the model. In this paper, we seek to address this challenge head-on. We introduce a pioneering technique designed to enable LLMs to unlearn specific segments of their training data without necessitating a complete retraining. Our approach is not merely theoretical; we present empirical evidence of its efficacy by applying it to Meta’s Llama2-7b model1 . As a proof of concept, we demonstrate that, while the original model can easily recover very detailed and nuanced information from the books, it’s possible for the model to essentially ”forget” the intricate narratives of the Harry Potter series ([Row07]), all while retaining its prowess on established benchmarks. To get a first impression of the fine-tuned model produced by our technique, Figure 1 compares the completions, on several prompts, of the baseline model (Llama2-7b-chat-hf) and a variant which has been fine-tuned for roughly 30 minutes on 4 A100-GPUs. Figure 2 compares the performance of these two models on some common benchmarks ([YBS19, CLC+19, ZHB+19, MCKS18, BHT+19, SLBBC19]) and Figure 3 compares the next token probability distributions for the sentence ”Harry Potter studies” over different steps of fine-tuning, showing how the most likely next token gradually shifts from ”magic” to generic completions. Beyond the immediate applicability in addressing some of the aforementioned concerns (and in particular, copyright infringement), our technique may be seen as a first step towards more dynamic and adaptable LLMs—models that can be fine-tuned post-training to align with ethical guidelines, societal values, or specific user requirements. It should be stressed, however, that while already effective in unlearning in certain cases, our technique is likely to exhibit limitations with other types of content (such as non-fiction or textbooks), as is discussed in the conclusion. Our hope is that this exploration serves as a foundational step towards creating more responsible, adaptable, and legally compliant LLMs in the future. 1.1 Related work While there is a growing body of work in the topic of unlearning in machine-learning in general (see [JLZ+22, NHN+22, ZNIS23] and references therein), the majority of works focus on classification tasks, while the literature concerning generative models or specifically LLMs is still quite slim. The very recent paper [ZFBH+23] highlights the related challenges and implications and discusses some high-level directions for potential mitigation. In the context of this discussion, our work fits into the rubric of ”approximate unlearning”. Recent works that propose concrete unlearning techniques for generative models are [JYY+22] which suggests a technique shown to address privacy risks in certain settings, and [WCY+23] which proposes an algorithm called knowledge-gap-alignment which may be in, certain cases, relevant for LLMs but relies on assumptions that do not seem to hold in our setting. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv 1 Our model can be found at https://huggingface.co/microsoft/Llama2-7b-WhoIsHarryPotter

Who’s Harry Potter? Approximate Unlearning in LLMs: Abstract and Introduction

Authors:
(1) Omer Rathore, Department of Physics, Durham University;
(2) Alastair Basden, Department of Physics, Durham University;
(3) Nicholas Chancellor, Department of Physics, Durham University and School of Computing, Newcastle University;
(4) Halim Kusumaatmaja, Department of Physics, Durham University and Institute for Multiscale Thermofluids, School of Engineering, The University of Edinburgh. Table of Links Abstract and Introduction Methods Quantum Annealing
Adaptive Mesh Refinement
Smoothed Particle Hydrodynamics Results Grid Based Application
Particle Based Application Conclusions and Outlook, Acknowledgements, and References Quantum Annealing Of the myriad possibilities[42] when it comes to implementing a quantum computer, currently the two leading paradigms are gate based quantum computing (QC)[43, 44] and adiabatic quantum computing (AQC)[45, 46]. In principle the two approaches have been shown[47] to be equivalent if allowing a polynomial overhead. Gate based QC can be thought of as the quantum analogue to classical computing where instead of logical gates acting on bits, the computation is performed by applying unitary gates to qubits. Alternatively, AQC relies on adiabatic time evolution of a quantum state from an initial, easy to prepare state to a final observed value as modelled by the Schrödinger equation. Moreover by operating under the umbrella of the adiabatic theorem, the protocol is guaranteed to obtain the ground state solution, at least in theory. However, in practice the requirements of true adiabacity and changing the Hamiltonian slowly enough can be excessively stringent. This is clearly evident in current architecture where thermal coupling with the environment is strong enough such that the timescales are of the order ∼ 10−100ns[53]. Furthermore the requirements on how quickly the Hamiltonian can be evolved (i.e. the annealing schedule) is dependant on the gap between the ground state and first excited solution, which is not usually known a priori. QA is thus a relaxation of these conditions that follows the same principles as AQC, but forms a distinctly separate algorithm, with the annealing schedule determined heuristically and strictly adiabatic conditions not guaranteed. The same time dependant Hamiltonian of Equation 1 is used to evolve the quantum state, however the system can no longer accurately be modeled by the Schrödinger equation, but would require master-equation based descriptions. As a result, QA forms a heuristic quantum algorithm that should be viewed as a statistical sampler rather than a deterministic solver, with the intended goal of maintaining a relatively high probability of remaining in the ground state or to at least be sufficiently close. The theory of quantum annealing is much less developed than AQC, partially due to the more complex setting. However, some progress has been made, for example in the topic of diabatic computing which considers rapid quenches far from the adiabatic limit[54]. In particular, in this regime a mechanism related to energy conservation provides an important guarantee on average optimality[55]. The expectation that Quantum Annealing (QA) might surpass classical algorithms is tied to the potential impact of quantum mechanical phenomena such as superposition, tunneling, and entanglement[31], [45] [57–59]. Fundamentally, quantum superposition and tunneling can enable transitions between states, even those separated by high energy barriers[60]. This suggests that a search algorithm utilizing QA could overcome local minima more easily by tunneling through these energy barriers as illustrated in Figure 1. QA is thus grounded in a model that more closely reflects the behavior of real quantum systems when compared to AQC. However, the potential for non-adiabatic effects complicates the distinction between its computational complexity and that of conventional computing. QA often involves the non-adiabatic evolution of mixed quantum states within an open system, and this approach is generally considered heuristic. Thus, it lacks the provable computational advantages that AQC provides[61]. The general approach to employing a quantum annealer shares many similarities across different architectures. This research, however, utilises D-Wave systems, which consist of superconducting qubits. Although there are several promising alternatives[62–64], these technologies are comparatively nascent in their development. Meanwhile D-Wave is the largest and most commonly used commercial QA platform currently and has been used extensively for both industry and research purposes. For a detailed, step-by-step guide on using a quantum annealer, interested readers are encouraged to consult the numerous introductory articles available on this topic[31], [61], [65]. In summary, the primary steps involved in utilizing an annealer include problem formulation, minor embedding, and sampling (i.e. the anneal itself). Problem formulation and how the classical simulations were run is described next, with the influence of embedding and sampling discussed as part of the results. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Omer Rathore, Department of Physics, Durham University; (2) Alastair Basden, Department of Physics, Durham University; (3) Nicholas Chancellor, Department of Physics, Durham University and School of Computing, Newcastle University; (4) Halim Kusumaatmaja, Department of Physics, Durham University and Institute for Multiscale Thermofluids, School of Engineering, The University of Edinburgh. Authors: Authors: (1) Omer Rathore, Department of Physics, Durham University; (2) Alastair Basden, Department of Physics, Durham University; (3) Nicholas Chancellor, Department of Physics, Durham University and School of Computing, Newcastle University; (4) Halim Kusumaatmaja, Department of Physics, Durham University and Institute for Multiscale Thermofluids, School of Engineering, The University of Edinburgh. Table of Links Abstract and Introduction Abstract and Introduction Abstract and Introduction Methods Methods Quantum Annealing Adaptive Mesh Refinement Smoothed Particle Hydrodynamics Quantum Annealing Quantum Annealing Adaptive Mesh Refinement Adaptive Mesh Refinement Smoothed Particle Hydrodynamics Smoothed Particle Hydrodynamics Results Results Grid Based Application Particle Based Application Grid Based Application Grid Based Application Particle Based Application Particle Based Application Conclusions and Outlook, Acknowledgements, and References Conclusions and Outlook, Acknowledgements, and References Conclusions and Outlook, Acknowledgements, and References Quantum Annealing Of the myriad possibilities[42] when it comes to implementing a quantum computer, currently the two leading paradigms are gate based quantum computing (QC)[43, 44] and adiabatic quantum computing (AQC)[45, 46]. In principle the two approaches have been shown[47] to be equivalent if allowing a polynomial overhead. Gate based QC can be thought of as the quantum analogue to classical computing where instead of logical gates acting on bits, the computation is performed by applying unitary gates to qubits. Alternatively, AQC relies on adiabatic time evolution of a quantum state from an initial, easy to prepare state to a final observed value as modelled by the Schrödinger equation. Moreover by operating under the umbrella of the adiabatic theorem, the protocol is guaranteed to obtain the ground state solution, at least in theory. However, in practice the requirements of true adiabacity and changing the Hamiltonian slowly enough can be excessively stringent. This is clearly evident in current architecture where thermal coupling with the environment is strong enough such that the timescales are of the order ∼ 10−100ns[53]. Furthermore the requirements on how quickly the Hamiltonian can be evolved (i.e. the annealing schedule) is dependant on the gap between the ground state and first excited solution, which is not usually known a priori. QA is thus a relaxation of these conditions that follows the same principles as AQC, but forms a distinctly separate algorithm, with the annealing schedule determined heuristically and strictly adiabatic conditions not guaranteed. The same time dependant Hamiltonian of Equation 1 is used to evolve the quantum state, however the system can no longer accurately be modeled by the Schrödinger equation, but would require master-equation based descriptions. As a result, QA forms a heuristic quantum algorithm that should be viewed as a statistical sampler rather than a deterministic solver, with the intended goal of maintaining a relatively high probability of remaining in the ground state or to at least be sufficiently close. The theory of quantum annealing is much less developed than AQC, partially due to the more complex setting. However, some progress has been made, for example in the topic of diabatic computing which considers rapid quenches far from the adiabatic limit[54]. In particular, in this regime a mechanism related to energy conservation provides an important guarantee on average optimality[55]. The expectation that Quantum Annealing (QA) might surpass classical algorithms is tied to the potential impact of quantum mechanical phenomena such as superposition, tunneling, and entanglement[31], [45] [57–59]. Fundamentally, quantum superposition and tunneling can enable transitions between states, even those separated by high energy barriers[60]. This suggests that a search algorithm utilizing QA could overcome local minima more easily by tunneling through these energy barriers as illustrated in Figure 1. QA is thus grounded in a model that more closely reflects the behavior of real quantum systems when compared to AQC. However, the potential for non-adiabatic effects complicates the distinction between its computational complexity and that of conventional computing. QA often involves the non-adiabatic evolution of mixed quantum states within an open system, and this approach is generally considered heuristic. Thus, it lacks the provable computational advantages that AQC provides[61]. The general approach to employing a quantum annealer shares many similarities across different architectures. This research, however, utilises D-Wave systems, which consist of superconducting qubits. Although there are several promising alternatives[62–64], these technologies are comparatively nascent in their development. Meanwhile D-Wave is the largest and most commonly used commercial QA platform currently and has been used extensively for both industry and research purposes. For a detailed, step-by-step guide on using a quantum annealer, interested readers are encouraged to consult the numerous introductory articles available on this topic[31], [61], [65]. In summary, the primary steps involved in utilizing an annealer include problem formulation, minor embedding, and sampling (i.e. the anneal itself). Problem formulation and how the classical simulations were run is described next, with the influence of embedding and sampling discussed as part of the results. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Load Balancing For High Performance Computing
Using Quantum Annealing: Quantum Annealing

Authors:
(1) Muneera Bano;
(2) Didar Zowghi;
(3) Vincenzo Gervasi;
(4) Rifat Shams. Table of Links Abstract, Impact Statement, and Introduction Defining Diversity and Inclusion in AI Research Motivation Research Methodology Results Discussion Conclusion and Future Work and References VI.DISCUSSION In this section we briefly discuss a few insights that are relevant in the wider context in which our proposed process for deriving D&I-focused user stories should take place, as well as discussing limitations of our research. A. Importance of Longterm Trade-off Analysis In the competitive AI-driven business landscape, prioritizing D&I over accuracy and profit may seem like a hard sell. However, it is essential to recognize that neglecting D&I can have adverse long-term effects on a company's reputation, including backlash, loss of user trust, and potential legal consequences. Recognizing that businesses may initially benefit from the higher accuracy and efficiency of AI systems that do not fully address D&I concerns, and may be tempted to save on related costs, the approach we have presented in this paper helps focus the investment on those facets of D&I that really matter in the specific context, by way of considering among the requirements only the user stories focusing on the attributes, values, roles and artifact that are relevant for the project. B. Responsibility of the AI development team While biased training datasets can perpetuate and reinforce discriminatory algorithms due to the use of shared and reusable datasets containing inherent biases [11] [31], a lack of diversity and inclusion in AI development teams can lead to unconscious biases and stereotypes, as team members may inadvertently project their own perceptions of reality or society into their work [32, 33]. C. D&I Requirements and Non-Functional Requirements Analysing D&I requirements in AI, like other nonfunctional requirements, entices concepts such as satisfaction up to a certain level (satisficing), inter-requirements structures based on positive and negative contributions to certain goals, and the determination of acceptable trade-offs in case of conflicts. By embodying general D&I guidelines into concrete user stories, our approach empowers developers to give them the same level of attention and rigour as other vital system characteristics. This approach ensures that AI systems are not only efficient and reliable but also ethically responsible and inclusive, catering to diverse user needs. D. Context and Culture The context surrounding AI system development plays a crucial role in implementing diversity and inclusion requirements. For example, Europe's General Data Protection Regulation (GDPR) [34] enforces strict privacy laws that organizations must adhere to when designing AI systems; this in turn may affect whether certain D&I requirements can be satisfied. In contrast, some countries with less individualistic cultures may have more hands-off privacy norms. The realization of D&I requirements in AI systems is also influenced by various factors such as organizational culture, governmental and legal frameworks, and societal norms that govern the development and deployment of these systems. E. Using GPT-4 for writing D&I in AI Requirements GPT-4, with its expansive training dataset, is well-equipped to assume diverse personas and extrapolate a broad range of requirements, reflecting potential challenges from myriad perspectives. This capacity aids in generating a holistic view of potential requirements, especially in the early stages of AI system design. However, while it can simulate multiple viewpoints, it inherently lacks the authenticity and depth of real human experiences, especially regarding nuanced aspects of diversity and inclusion. The subjective intricacies and tangible challenges faced by individuals in their lived experiences often provide invaluable insights that may not be entirely understood by GPT-4. Thus, combining the broad, simulated perspectives of GPT-4 with the grounded, real-world experiences of human stakeholders can offer a synergistic approach. Together, they provide a richer, more comprehensive information base that can immensely benefit analysts and developers, ensuring AI systems are both inclusive and cognizant of the intricate facets of human diversity. F. Limitations Our research, despite employing a rigorous evidence-based methodology to identify D&I requirement themes, has certain limitations in coverage and applicability. One limitation is that the themes are derived solely from published and publicly available research, which might not cover the entire spectrum of D&I issues in AI systems. Additionally, although the identified requirements are generalizable and applicable to a broad range of AI systems, further analysisis necessary to tailor and adapt them to the specific context of an individual AI system project. The process of deriving D&I user stories that we have outlined is highly reliant not only on the requirements engineer’s professional ability but also on their ethical and social sensibility. We believe this to be necessary and the only way to give due consideration to the nuances of ethical reasoning. These limitations imply that the current research serves as a foundation for understanding D&I requirementsin AI, but there is a need for more in-depth examination and customization to address unique challenges and peculiarities that may arise in specific AI implementations and social contexts. This paper is available on arxiv under CC 4.0 license. Authors: (1) Muneera Bano; (2) Didar Zowghi; (3) Vincenzo Gervasi; (4) Rifat Shams. Authors: Authors: (1) Muneera Bano; (2) Didar Zowghi; (3) Vincenzo Gervasi; (4) Rifat Shams. Table of Links Abstract, Impact Statement, and Introduction Abstract, Impact Statement, and Introduction Defining Diversity and Inclusion in AI Defining Diversity and Inclusion in AI Research Motivation Research Motivation Research Methodology Research Methodology Results Results Discussion Discussion Conclusion and Future Work and References Conclusion and Future Work and References VI.DISCUSSION In this section we briefly discuss a few insights that are relevant in the wider context in which our proposed process for deriving D&I-focused user stories should take place, as well as discussing limitations of our research. A. Importance of Longterm Trade-off Analysis A. Importance of Longterm Trade-off Analysis In the competitive AI-driven business landscape, prioritizing D&I over accuracy and profit may seem like a hard sell. However, it is essential to recognize that neglecting D&I can have adverse long-term effects on a company's reputation, including backlash, loss of user trust, and potential legal consequences. Recognizing that businesses may initially benefit from the higher accuracy and efficiency of AI systems that do not fully address D&I concerns, and may be tempted to save on related costs, the approach we have presented in this paper helps focus the investment on those facets of D&I that really matter in the specific context, by way of considering among the requirements only the user stories focusing on the attributes, values, roles and artifact that are relevant for the project. B. Responsibility of the AI development team B. Responsibility of the AI development team While biased training datasets can perpetuate and reinforce discriminatory algorithms due to the use of shared and reusable datasets containing inherent biases [11] [31], a lack of diversity and inclusion in AI development teams can lead to unconscious biases and stereotypes, as team members may inadvertently project their own perceptions of reality or society into their work [32, 33]. C. D&I Requirements and Non-Functional Requirements C. D&I Requirements and Non-Functional Requirements Analysing D&I requirements in AI, like other nonfunctional requirements, entices concepts such as satisfaction up to a certain level (satisficing), inter-requirements structures based on positive and negative contributions to certain goals, and the determination of acceptable trade-offs in case of conflicts. By embodying general D&I guidelines into concrete user stories, our approach empowers developers to give them the same level of attention and rigour as other vital system characteristics. This approach ensures that AI systems are not only efficient and reliable but also ethically responsible and inclusive, catering to diverse user needs. D. Context and Culture D. Context and Culture The context surrounding AI system development plays a crucial role in implementing diversity and inclusion requirements. For example, Europe's General Data Protection Regulation (GDPR) [34] enforces strict privacy laws that organizations must adhere to when designing AI systems; this in turn may affect whether certain D&I requirements can be satisfied. In contrast, some countries with less individualistic cultures may have more hands-off privacy norms. The realization of D&I requirements in AI systems is also influenced by various factors such as organizational culture, governmental and legal frameworks, and societal norms that govern the development and deployment of these systems. E. Using GPT-4 for writing D&I in AI Requirements E. Using GPT-4 for writing D&I in AI Requirements GPT-4, with its expansive training dataset, is well-equipped to assume diverse personas and extrapolate a broad range of requirements, reflecting potential challenges from myriad perspectives. This capacity aids in generating a holistic view of potential requirements, especially in the early stages of AI system design. However, while it can simulate multiple viewpoints, it inherently lacks the authenticity and depth of real human experiences, especially regarding nuanced aspects of diversity and inclusion. The subjective intricacies and tangible challenges faced by individuals in their lived experiences often provide invaluable insights that may not be entirely understood by GPT-4. Thus, combining the broad, simulated perspectives of GPT-4 with the grounded, real-world experiences of human stakeholders can offer a synergistic approach. Together, they provide a richer, more comprehensive information base that can immensely benefit analysts and developers, ensuring AI systems are both inclusive and cognizant of the intricate facets of human diversity. F. Limitations F. Limitations Our research, despite employing a rigorous evidence-based methodology to identify D&I requirement themes, has certain limitations in coverage and applicability. One limitation is that the themes are derived solely from published and publicly available research, which might not cover the entire spectrum of D&I issues in AI systems. Additionally, although the identified requirements are generalizable and applicable to a broad range of AI systems, further analysisis necessary to tailor and adapt them to the specific context of an individual AI system project. The process of deriving D&I user stories that we have outlined is highly reliant not only on the requirements engineer’s professional ability but also on their ethical and social sensibility. We believe this to be necessary and the only way to give due consideration to the nuances of ethical reasoning. These limitations imply that the current research serves as a foundation for understanding D&I requirementsin AI, but there is a need for more in-depth examination and customization to address unique challenges and peculiarities that may arise in specific AI implementations and social contexts. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

The Hidden Costs of Neglecting Diversity Inclusion in AI Systems

Secrets management is not just about safeguarding sensitive information; it is about ensuring seamless and secure access to this information whenever and wherever it's needed. Leaked secrets are the root cause of many prominent security breaches so we talked to Vlad Matsiiako from Infisical to learn more about secret management and why it’s still a big problem for many companies. Here is the full video interview: https://youtu.be/z0NC89nZJrM What is Secrets Management? Secrets management involves the secure handling of sensitive information, such as API keys, passwords, certificates, and access tokens. These secrets are necessary for various parts of an infrastructure to interact securely. For instance, databases need to connect to applications, and servers need to communicate with each other. Secrets management solutions ensure that these interactions happen securely and efficiently, without exposing sensitive information to unauthorized parties. What is the Problem? As software systems become more complex, the challenge of securely managing and distributing secrets intensifies. Different parts of a system—whether they are databases, applications, or servers—need to be securely interconnected. The traditional methods of managing these secrets are often manual, prone to errors, and not scalable. This can lead to significant security vulnerabilities, such as secret leaks or unauthorized access, which can have severe repercussions, including data breaches and compliance issues. Different Approaches There are several approaches to secrets management: Manual Management: Some companies opt to manage secrets manually, using secure storage solutions like Key Management Systems (KMS). While this method can be straightforward, it often lacks automation and scalability, leading to potential security gaps.


Automated Solutions: Tools like Terraform Vault have become popular, offering automated secrets management. These tools provide a more structured and scalable way to manage secrets but can be complex to implement and maintain.


Dynamic Secrets: More advanced solutions offer dynamic secrets management, where secrets are generated on-the-fly and are valid for a limited time or a specific number of requests. This approach minimizes the risk of secret leaks and unauthorized access.


Integrated Solutions: Platforms like Infisical offer integrated secrets management solutions that provide seamless access control, secret rotation, and automation across different environments (development, staging, production). Types of Companies, When They Adopt a Solution, and Their Motivations The adoption of secrets management solutions varies based on the company's size, industry, and stage of development: Early-Stage Startups: These companies often adopt secrets management solutions early, especially if they operate in highly regulated industries like finance or healthcare. The motivation here is to establish strong security foundations from the beginning.


Growth-Stage Companies: Companies at this stage (Series B and beyond) adopt secrets management solutions to improve operational efficiency and enhance their security posture. As their infrastructure becomes more complex, managing secrets manually becomes impractical.


Large Enterprises and Governments: These organizations have strict security requirements and often have specific policies on how secrets should be managed. Their motivation is not only security but also compliance with regulatory standards. Security Security is the primary driver behind secrets management. Effective secrets management minimizes the risk of unauthorized access and data breaches. It ensures that sensitive information is only accessible to those who need it, and it is done so in a secure and controlled manner. Implementing features like automatic secret rotation and dynamic secret generation further enhances security by reducing the attack surface and limiting the potential damage of a compromised secret. Why Isn’t This Solved Already? Despite the existence of secrets management solutions, many companies still experience secret leaks and security breaches. This is often due to the complexity of these solutions, which can be challenging for developers to implement and use effectively. If a secrets management tool is overly complicated, developers might bypass it, leading to security vulnerabilities. Moreover, the dynamic nature of modern infrastructures, with multiple environments and a myriad of integration points, adds to the complexity. At Infisical, the focus is on making secrets management more accessible and user-friendly for developers, ensuring that security practices are embedded into their workflows without adding unnecessary friction. By providing robust tools and seamless integrations, Infisical aims to reduce the complexity and enhance the security of secrets management. Conclusion Secrets management is a critical aspect of modern cybersecurity, ensuring that sensitive information is handled securely and efficiently across different environments. While there are various approaches and solutions available, the key to effective secrets management lies in balancing security with ease of use. By understanding the unique needs and motivations of different types of companies, we can better tailor secrets management solutions to address their specific challenges and enhance their overall security posture. Secrets management is not just about safeguarding sensitive information; it is about ensuring seamless and secure access to this information whenever and wherever it's needed. Leaked secrets are the root cause of many prominent security breaches so we talked to Vlad Matsiiako from Infisical to learn more about secret management and why it’s still a big problem for many companies. Infisical Here is the full video interview: https://youtu.be/z0NC89nZJrM https://youtu.be/z0NC89nZJrM What is Secrets Management? Secrets management involves the secure handling of sensitive information, such as API keys, passwords, certificates, and access tokens. These secrets are necessary for various parts of an infrastructure to interact securely. For instance, databases need to connect to applications, and servers need to communicate with each other. Secrets management solutions ensure that these interactions happen securely and efficiently, without exposing sensitive information to unauthorized parties. What is the Problem? As software systems become more complex, the challenge of securely managing and distributing secrets intensifies. Different parts of a system—whether they are databases, applications, or servers—need to be securely interconnected. The traditional methods of managing these secrets are often manual, prone to errors, and not scalable. This can lead to significant security vulnerabilities, such as secret leaks or unauthorized access, which can have severe repercussions, including data breaches and compliance issues. Different Approaches There are several approaches to secrets management: Manual Management: Some companies opt to manage secrets manually, using secure storage solutions like Key Management Systems (KMS). While this method can be straightforward, it often lacks automation and scalability, leading to potential security gaps. Automated Solutions: Tools like Terraform Vault have become popular, offering automated secrets management. These tools provide a more structured and scalable way to manage secrets but can be complex to implement and maintain. Dynamic Secrets: More advanced solutions offer dynamic secrets management, where secrets are generated on-the-fly and are valid for a limited time or a specific number of requests. This approach minimizes the risk of secret leaks and unauthorized access. Integrated Solutions: Platforms like Infisical offer integrated secrets management solutions that provide seamless access control, secret rotation, and automation across different environments (development, staging, production). Manual Management: Some companies opt to manage secrets manually, using secure storage solutions like Key Management Systems (KMS). While this method can be straightforward, it often lacks automation and scalability, leading to potential security gaps. Manual Management: Some companies opt to manage secrets manually, using secure storage solutions like Key Management Systems (KMS). While this method can be straightforward, it often lacks automation and scalability, leading to potential security gaps. Automated Solutions: Tools like Terraform Vault have become popular, offering automated secrets management. These tools provide a more structured and scalable way to manage secrets but can be complex to implement and maintain. Automated Solutions: Tools like Terraform Vault have become popular, offering automated secrets management. These tools provide a more structured and scalable way to manage secrets but can be complex to implement and maintain. Dynamic Secrets: More advanced solutions offer dynamic secrets management, where secrets are generated on-the-fly and are valid for a limited time or a specific number of requests. This approach minimizes the risk of secret leaks and unauthorized access. Dynamic Secrets: More advanced solutions offer dynamic secrets management, where secrets are generated on-the-fly and are valid for a limited time or a specific number of requests. This approach minimizes the risk of secret leaks and unauthorized access. Integrated Solutions: Platforms like Infisical offer integrated secrets management solutions that provide seamless access control, secret rotation, and automation across different environments (development, staging, production). Integrated Solutions: Platforms like Infisical offer integrated secrets management solutions that provide seamless access control, secret rotation, and automation across different environments (development, staging, production). Types of Companies, When They Adopt a Solution, and Their Motivations The adoption of secrets management solutions varies based on the company's size, industry, and stage of development: Early-Stage Startups: These companies often adopt secrets management solutions early, especially if they operate in highly regulated industries like finance or healthcare. The motivation here is to establish strong security foundations from the beginning. Growth-Stage Companies: Companies at this stage (Series B and beyond) adopt secrets management solutions to improve operational efficiency and enhance their security posture. As their infrastructure becomes more complex, managing secrets manually becomes impractical. Large Enterprises and Governments: These organizations have strict security requirements and often have specific policies on how secrets should be managed. Their motivation is not only security but also compliance with regulatory standards. Early-Stage Startups: These companies often adopt secrets management solutions early, especially if they operate in highly regulated industries like finance or healthcare. The motivation here is to establish strong security foundations from the beginning. Early-Stage Startups: These companies often adopt secrets management solutions early, especially if they operate in highly regulated industries like finance or healthcare. The motivation here is to establish strong security foundations from the beginning. Growth-Stage Companies: Companies at this stage (Series B and beyond) adopt secrets management solutions to improve operational efficiency and enhance their security posture. As their infrastructure becomes more complex, managing secrets manually becomes impractical. Growth-Stage Companies: Companies at this stage (Series B and beyond) adopt secrets management solutions to improve operational efficiency and enhance their security posture. As their infrastructure becomes more complex, managing secrets manually becomes impractical. Large Enterprises and Governments: These organizations have strict security requirements and often have specific policies on how secrets should be managed. Their motivation is not only security but also compliance with regulatory standards. Large Enterprises and Governments: These organizations have strict security requirements and often have specific policies on how secrets should be managed. Their motivation is not only security but also compliance with regulatory standards. Security Security is the primary driver behind secrets management. Effective secrets management minimizes the risk of unauthorized access and data breaches. It ensures that sensitive information is only accessible to those who need it, and it is done so in a secure and controlled manner. Implementing features like automatic secret rotation and dynamic secret generation further enhances security by reducing the attack surface and limiting the potential damage of a compromised secret. Why Isn’t This Solved Already? Despite the existence of secrets management solutions, many companies still experience secret leaks and security breaches. This is often due to the complexity of these solutions, which can be challenging for developers to implement and use effectively. If a secrets management tool is overly complicated, developers might bypass it, leading to security vulnerabilities. Moreover, the dynamic nature of modern infrastructures, with multiple environments and a myriad of integration points, adds to the complexity. At Infisical, the focus is on making secrets management more accessible and user-friendly for developers, ensuring that security practices are embedded into their workflows without adding unnecessary friction. By providing robust tools and seamless integrations, Infisical aims to reduce the complexity and enhance the security of secrets management. Conclusion Secrets management is a critical aspect of modern cybersecurity, ensuring that sensitive information is handled securely and efficiently across different environments. While there are various approaches and solutions available, the key to effective secrets management lies in balancing security with ease of use. By understanding the unique needs and motivations of different types of companies, we can better tailor secrets management solutions to address their specific challenges and enhance their overall security posture.

Shhh...The Secret to Secrets Management: What You Need to Know

We all want love. The problem is, how does one find it? For thousands of years, humans relied on personal attraction, matchmakers, dating, and myriad other ways to attract a mate. Today, many people use apps like Tinder, Bumble, or Hinge for example. Dating apps are the best and worst thing that has happened to mankind. On the one hand, it has given us the ability to find thousands of people with no effort. On the other hand, the apps have made us lazy, picky, and too discerning. There may be no going back as we now have millions of potential partners at our fingertips and are not just limited to our little village. But what if the apps are taken to a new level and incorporate AI? While AI can be used for nefarious purposes such as generating fake profile text or creating fake profile images, AI is used for several great reasons to help us find love. For instance, AI can help personalize our experience and narrow down choices to people we would most likely want to meet. It can also help introverted people start a conversation. One of the main barriers to finding love is language. If we can only meet people who speak our language, we are cut off from potential partners who only speak a different language. But what if AI can now help us, through a multi-lingual avatar, speak to anyone we want around the world? Nvidia-backed Generative AI firm Synthesia for example has introduced new software that allows companies to make multilingual video presentations using just your phone or webcam. Taking this one step further, the technology can be personalized for individual users. In its recent keynote event introducing the software, one of the biggest new features the firm showed off was an option to make AI-generated avatars by recording footage using a webcam or your phone. The firm demonstrated how users can also clone their voice to have the avatars speak in multiple different languages. This could be perfect for dating. As long as users adhere to the truth and there are vetting standards in place like there are today, people can meet anyone around the world and converse in any language. Of course, for now this only works through technology and meeting in person and speaking in two different languages will remain difficult until technology improves, and auto-translation apps do a better job at recording and translating. At the moment there are apps such as Google Translate which can be used to listen to a recording in one language and then translate it into any number of other languages. This can be used for conversation but only in a halting and disruptive way that blocks the natural flow of conversation. The question is whether we can develop AI technology that acts a bit like the character Janet from that Netflix show “The Good Place”, where she becomes available just by mentioning her name and helps with any information you need. While still a far way off, the ability to speak to anyone and understand them while conversing in two different languages is drawing closer and, thanks to AI, the language barrier will become history. We all want love. The problem is, how does one find it? For thousands of years, humans relied on personal attraction, matchmakers, dating, and myriad other ways to attract a mate. Today, many people use apps like Tinder , Bumble, or Hinge for example. Tinder Dating apps are the best and worst thing that has happened to mankind. On the one hand, it has given us the ability to find thousands of people with no effort. On the other hand, the apps have made us lazy, picky, and too discerning. There may be no going back as we now have millions of potential partners at our fingertips and are not just limited to our little village. But what if the apps are taken to a new level and incorporate AI? While AI can be used for nefarious purposes such as generating fake profile text or creating fake profile images, AI is used for several great reasons to help us find love. several great reasons For instance, AI can help personalize our experience and narrow down choices to people we would most likely want to meet. It can also help introverted people start a conversation. One of the main barriers to finding love is language. If we can only meet people who speak our language, we are cut off from potential partners who only speak a different language. But what if AI can now help us, through a multi-lingual avatar, speak to anyone we want around the world? Nvidia-backed Generative AI firm Synthesia for example has introduced new software that allows companies to make multilingual video presentations using just your phone or webcam. introduced new software Taking this one step further, the technology can be personalized for individual users. In its recent keynote event introducing the software, one of the biggest new features the firm showed off was an option to make AI-generated avatars by recording footage using a webcam or your phone. The firm demonstrated how users can also clone their voice to have the avatars speak in multiple different languages. This could be perfect for dating. As long as users adhere to the truth and there are vetting standards in place like there are today, people can meet anyone around the world and converse in any language. Of course, for now this only works through technology and meeting in person and speaking in two different languages will remain difficult until technology improves, and auto-translation apps do a better job at recording and translating. At the moment there are apps such as Google Translate which can be used to listen to a recording in one language and then translate it into any number of other languages. This can be used for conversation but only in a halting and disruptive way that blocks the natural flow of conversation. The question is whether we can develop AI technology that acts a bit like the character Janet from that Netflix show “The Good Place”, where she becomes available just by mentioning her name and helps with any information you need. Janet While still a far way off, the ability to speak to anyone and understand them while conversing in two different languages is drawing closer and, thanks to AI, the language barrier will become history.

It’s Time for AI To Help Us Find Love

Effortlessly Launch LangChain APIs with LangServe and MinIO Integration

অনায়াসে LangServe এবং MinIO ইন্টিগ্রেশন সহ LangChain API চালু করুন

Müheloses Starten von LangChain-APIs mit LangServe- und MinIO-Integration

Lance las API de LangChain sin esfuerzo con la integración de LangServe y MinIO

Lancez sans effort les API LangChain avec l'intégration LangServe et MinIO

LangServe और MinIO एकीकरण के साथ LangChain API को आसानी से लॉन्च करें

LangServe と MinIO の統合により LangChain API を簡単に起動

LangServe 및 MinIO 통합을 통해 LangChain API를 손쉽게 실행

Inicie APIs LangChain sem esforço com integração LangServe e MinIO

Легко запускайте API-интерфейсы LangChain с интеграцией LangServe и MinIO.

LangServe ve MinIO Entegrasyonu ile LangChain API'lerini Zahmetsizce Başlatın

Dễ dàng khởi chạy API LangChain với tích hợp LangServe và MinIO

通过 LangServe 和 MinIO 集成轻松启动 LangChain API

Authors:
(1) Kenan Begovic, currently a Ph.D. candidate in Computer Science at Qatar University. He received his MS in Information and Computer Security from University of Liverpool;
(2) Abdulaziz Al-Ali ,received the Ph.D. degree in machine learning from the University of Miami, FL, USA, in 2016 and he is currently an Assistant Professor in the Computer Science and Engineering Department, and director of the KINDI Center for Computing Research at Qatar University;
(3) Qutaibah Malluhi, a Professor at the Department of Computer Science and Engineering at Qatar University (QU). Table of Links Abstract and 1 Introduction
2. On crypto-ransomware behavior and methodology
3. Detection of encryption
4. Conclusion
Declaration of Competing Interest, CRediT authorship contribution statement, Data availability, and References 4. Conclusion In closing, our journey through the myriad proposals relating to the detection of pre-encryption and encryption activities by cryptographic ransomware has surfaced some findings and lessons learned. The dynamic nature of ransomware necessitates constant vigilance, innovation, and adaptation. Each of the surveyed methodologies has its own advantages and disadvantages. This all is the reason we present section 4.1, Lessons learned. Our findings in section 4.2 underscore the importance of multi-layered, robust detection mechanisms and the need for more research focusing on encryption as the major motivation for the attack. We hope this analysis serves as a catalyst for further advancements and a guidepost for future endeavors in combating cryptographic ransomware. 4.1. Lessons learned In this paper, we reviewed cryptographic ransomware from the perspective of what we believe is its differentiating factor from other families of cyberattacks, namely encryption. Here we summarize the lessons learned in this survey. Foremost, cryptoransomware is an increasing threat that cripples critical capabilities for both public and commercial services for an extended period of time. When we add the amounts of paid ransoms, losses are in the tens of millions of dollars. They could result in loss of human life, taking into consideration ransomware groups’ “taste” for medical facilities during the time of the COVID-19 pandemic, or even involve themselves in cyber warfare. Secondly, researchers have seen a discrepancy in describing and categorizing ransomware ranging from plain malware to sophisticated cyber kill chains representing the activity of sophisticated APT-like threat actors. We observed ransomware through our cyber kill chain. Thirdly, from a research perspective, we have seen many proposals dealing with different aspects of crypto-ransomware, and most of them take a hybrid approach to deliver the solution. Fourthly, we have seen that despite the academic community’s mature and focused research, commercial solutions mainly apply machine learning and anomaly detection solutions. Finally, we have seen that specialized research around the problem of encryption detection and general control of encryption operations is in the apparent minority among the research topics into crypto-ransomware. Overall, there is no one-size-fits-all answer to the most effective approach. The effectiveness of each method depends on the specific ransomware and its behavior. Combining all approaches and machine learning may offer the best defense against ransomware attacks. Adding to that behavioral analysis and user awareness, effectiveness increases significantly. The information in Table 3 summarizes the pros and cons of each methodology, providing the argument for the hybrid approach advantage. These conclusions result from a review and study of different proposals presented in this survey and our research into commercial solutions that offer capabilities that could be used to detect pre-encryption and encryption activities by ransomware attacks. 4.2. Main findings We focused on the Encryption phase described in our cyber kill chain and divided various methodologies into three major groups: - API and system call monitoring-based detection - I/O monitoring-based detection - file system monitoring-based detection Reviewing the work of researchers through the prism of those three detection methodologies, it was not surprising that, in the case of complete proposals for detection, researchers preferred a hybrid approach utilizing primarily combinations of the three or fewer methodologies, with machine learning being the preferred tool in many proposals. Most of the research was focused on Windows operating systems, but Android mobile operating systems and a variety of Internet of Things applications were also present in the reviewed work. The overall conclusion of this survey is that more methods and techniques described in the surveyed research efforts should be utilized in real-life products by trying to remove some of the common obstacles already described. Detecting pre-encryption and encryption activities provides a high level of confidence that ransomware can be intercepted before doing severe damage. Introducing some benchmark methods in the detection of encryption along with appropriate datasets is a desperate need in this area of research. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Kenan Begovic, currently a Ph.D. candidate in Computer Science at Qatar University. He received his MS in Information and Computer Security from University of Liverpool; (2) Abdulaziz Al-Ali ,received the Ph.D. degree in machine learning from the University of Miami, FL, USA, in 2016 and he is currently an Assistant Professor in the Computer Science and Engineering Department, and director of the KINDI Center for Computing Research at Qatar University; (3) Qutaibah Malluhi, a Professor at the Department of Computer Science and Engineering at Qatar University (QU). Authors: Authors: (1) Kenan Begovic, currently a Ph.D. candidate in Computer Science at Qatar University. He received his MS in Information and Computer Security from University of Liverpool; (2) Abdulaziz Al-Ali ,received the Ph.D. degree in machine learning from the University of Miami, FL, USA, in 2016 and he is currently an Assistant Professor in the Computer Science and Engineering Department, and director of the KINDI Center for Computing Research at Qatar University; (3) Qutaibah Malluhi, a Professor at the Department of Computer Science and Engineering at Qatar University (QU). Table of Links Abstract and 1 Introduction 2. On crypto-ransomware behavior and methodology 3. Detection of encryption 4. Conclusion Declaration of Competing Interest, CRediT authorship contribution statement, Data availability, and References Abstract and 1 Introduction Abstract and 1 Introduction 2. On crypto-ransomware behavior and methodology 2. On crypto-ransomware behavior and methodology 3. Detection of encryption 3. Detection of encryption 4. Conclusion 4. Conclusion Declaration of Competing Interest, CRediT authorship contribution statement, Data availability, and References Declaration of Competing Interest, CRediT authorship contribution statement, Data availability, and References 4. Conclusion In closing, our journey through the myriad proposals relating to the detection of pre-encryption and encryption activities by cryptographic ransomware has surfaced some findings and lessons learned. The dynamic nature of ransomware necessitates constant vigilance, innovation, and adaptation. Each of the surveyed methodologies has its own advantages and disadvantages. This all is the reason we present section 4.1, Lessons learned. Our findings in section 4.2 underscore the importance of multi-layered, robust detection mechanisms and the need for more research focusing on encryption as the major motivation for the attack. We hope this analysis serves as a catalyst for further advancements and a guidepost for future endeavors in combating cryptographic ransomware. 4.1. Lessons learned 4.1. Lessons learned In this paper, we reviewed cryptographic ransomware from the perspective of what we believe is its differentiating factor from other families of cyberattacks, namely encryption. Here we summarize the lessons learned in this survey. Foremost, cryptoransomware is an increasing threat that cripples critical capabilities for both public and commercial services for an extended period of time. When we add the amounts of paid ransoms, losses are in the tens of millions of dollars. They could result in loss of human life, taking into consideration ransomware groups’ “taste” for medical facilities during the time of the COVID-19 pandemic, or even involve themselves in cyber warfare. Secondly, researchers have seen a discrepancy in describing and categorizing ransomware ranging from plain malware to sophisticated cyber kill chains representing the activity of sophisticated APT-like threat actors. We observed ransomware through our cyber kill chain. Thirdly, from a research perspective, we have seen many proposals dealing with different aspects of crypto-ransomware, and most of them take a hybrid approach to deliver the solution. Fourthly, we have seen that despite the academic community’s mature and focused research, commercial solutions mainly apply machine learning and anomaly detection solutions. Finally, we have seen that specialized research around the problem of encryption detection and general control of encryption operations is in the apparent minority among the research topics into crypto-ransomware. Overall, there is no one-size-fits-all answer to the most effective approach. The effectiveness of each method depends on the specific ransomware and its behavior. Combining all approaches and machine learning may offer the best defense against ransomware attacks. Adding to that behavioral analysis and user awareness, effectiveness increases significantly. The information in Table 3 summarizes the pros and cons of each methodology, providing the argument for the hybrid approach advantage. These conclusions result from a review and study of different proposals presented in this survey and our research into commercial solutions that offer capabilities that could be used to detect pre-encryption and encryption activities by ransomware attacks. 4.2. Main findings 4.2. Main findings We focused on the Encryption phase described in our cyber kill chain and divided various methodologies into three major groups: - API and system call monitoring-based detection - I/O monitoring-based detection - file system monitoring-based detection Reviewing the work of researchers through the prism of those three detection methodologies, it was not surprising that, in the case of complete proposals for detection, researchers preferred a hybrid approach utilizing primarily combinations of the three or fewer methodologies, with machine learning being the preferred tool in many proposals. Most of the research was focused on Windows operating systems, but Android mobile operating systems and a variety of Internet of Things applications were also present in the reviewed work. The overall conclusion of this survey is that more methods and techniques described in the surveyed research efforts should be utilized in real-life products by trying to remove some of the common obstacles already described. Detecting pre-encryption and encryption activities provides a high level of confidence that ransomware can be intercepted before doing severe damage. Introducing some benchmark methods in the detection of encryption along with appropriate datasets is a desperate need in this area of research. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Cryptographic Ransomware Encryption Detection: Survey: Conclusion

Now that the mitigation measures of COVID-19 have made a deep stamp upon the global economy, many who were laid off or lost businesses during the various lockdowns of 2020 and 2021 are starting to renew their economic earnings through what is euphemistically called the “gig economy,” a term that grown in popularity over the past decade. Still, this term has taken on new proportions in the post-pandemic United States. Now, freelancing jobs offered through apps like Uber, Lyft, Doordash, Deliveroo, and many other app-based services, are being taken on by former professionals During the Covid pandemic, newer, digital-first work platforms like Uber Eats surged in popularity as millions were forced to shelter at home. Those who could afford it turned to online delivery apps. Scientific data indicates that the number of people working with these platforms has grown, too. “The Evolution of Platform Gig Work, 2012-2021” published by the University of Chicago in May 2023, tracked the earnings of gig workers through tax filings. What this study uncovered was that the number of people who report income from platform-based gig work to the IRS has exploded in recent years from just over one million to nearly five million workers. Aside from being a clear indication of just how many more people use tech platforms to help earn a living, this demonstrates how the gig economy is rising in popularity even for the convenience of those who prefer to work part-time or a specific schedule in order to take on other caregiving roles at home. But the gig economy is not limited to working-class delivery jobs. There has been a longstanding gig economy among the educated for decades within higher education—namely that of adjunct faculty. These professors are often graduate students working on their PhD, but even today, more and more are PhDs with no hope for full-time employment in sight. Adjunct faculty are not only paid a small amount compared to the tenure-track cohorts, but they lack job security, benefits, and even—depending on the institution—health insurance and sick leave. While evidence abounds as to how the global pandemic pushed the gig economy into popularity, even by its previous critics, the necessity of survival was likely the incentive to enter into this job market. With no possibility of leaving the house for millions, online platforms became the obvious solution, to include the class of people who decided to enter online trading from their homes. The impact of COVID-19 on global economies made a dent even among the middle classes that were willing to take marked investment risks from their sofas where Forex trading and myriad investment platforms became the focus for a number of working populations—especially the younger generations with savings, yet ample time to participate in the share and currency markets. What is surprising about the post-pandemic gig economy is not so much its enduring popularity, but how these jobs have alerted workers far beyond the gig theatre, that perhaps they too ought to have a greater say in their working hours, the ability to care for children and elderly relatives, and most of all, it has set off a larger conversation about the four-day work week, something currently under experimentation in Mexico, the Dominican Republic and the United Kingdom, among many countries. Perhaps the gig economy’s greatest legacy might actually be that it reshapes the landscape of labor rights—from the blue to the white-collar workers—across the planet. Now that the mitigation measures of COVID-19 have made a deep stamp upon the global economy, many who were laid off or lost businesses during the various lockdowns of 2020 and 2021 are starting to renew their economic earnings through what is euphemistically called the “gig economy,” a term that grown in popularity over the past decade. Still, this term has taken on new proportions in the post-pandemic United States. Now, freelancing jobs offered through apps like Uber, Lyft, Doordash, Deliveroo, and many other app-based services, are being taken on by former professionals During the Covid pandemic, newer, digital-first work platforms like Uber Eats surged in popularity as millions were forced to shelter at home. Those who could afford it turned to online delivery apps. Scientific data indicates that the number of people working with these platforms has grown, too. “ The Evolution of Platform Gig Work, 2012-2021 ” published by the University of Chicago in May 2023, tracked the earnings of gig workers through tax filings. What this study uncovered was that the number of people who report income from platform-based gig work to the IRS has exploded in recent years from just over one million to nearly five million workers. The Evolution of Platform Gig Work, 2012-2021 The Evolution of Platform Gig Work, 2012-2021 Aside from being a clear indication of just how many more people use tech platforms to help earn a living, this demonstrates how the gig economy is rising in popularity even for the convenience of those who prefer to work part-time or a specific schedule in order to take on other caregiving roles at home. But the gig economy is not limited to working-class delivery jobs. There has been a longstanding gig economy among the educated for decades within higher education—namely that of adjunct faculty . These professors are often graduate students working on their PhD, but even today, more and more are PhDs with no hope for full-time employment in sight. adjunct faculty adjunct faculty Adjunct faculty are not only paid a small amount compared to the tenure-track cohorts, but they lack job security, benefits, and even—depending on the institution—health insurance and sick leave. While evidence abounds as to how the global pandemic pushed the gig economy into popularity, even by its previous critics, the necessity of survival was likely the incentive to enter into this job market. With no possibility of leaving the house for millions, online platforms became the obvious solution, to include the class of people who decided to enter online trading from their homes. evidence abounds evidence abounds The impact of COVID-19 on global economies made a dent even among the middle classes that were willing to take marked investment risks from their sofas where Forex trading and myriad investment platforms became the focus for a number of working populations—especially the younger generations with savings, yet ample time to participate in the share and currency markets. marked investment risks marked investment risks Forex trading Forex trading investment platforms younger generations What is surprising about the post-pandemic gig economy is not so much its enduring popularity, but how these jobs have alerted workers far beyond the gig theatre, that perhaps they too ought to have a greater say in their working hours, the ability to care for children and elderly relatives, and most of all, it has set off a larger conversation about the four-day work week, something currently under experimentation in Mexico, the Dominican Republic and the United Kingdom, among many countries. Perhaps the gig economy’s greatest legacy might actually be that it reshapes the landscape of labor rights—from the blue to the white-collar workers—across the planet.

The Rise and Rise of the Gig Economy in a Post-Pandemic World

You probably like sandwiches as a snack, but what about being sandwiched by your competitors on Google Search? Elastic Email (a Mailchimp competitor) has been advertising on Google for months. Google Ads constitute ~40% of their ad spend, so they must get fantastic results there, right? WRONG!! If you Google Elastic Email right now, here’s what you’ll see: They. Got. Sandwiched! If you went ‘oooooof’, then you’re a marketer at heart and should read on. Google Likes Your Money - Not You. When Google opened up the bidding mechanism for ADs, the first casualty was Ethics. Until then, Marketers refrained from encroaching upon competitors’ branding, and any harmless banter was considered' a bold move’—remember the Cola Wars? But today, we’re all cutthroats over that $5 CPM. Thanks, Google! Sandwiches like the one above are the norm, not the exception, while Google laughs its way to the bank. So, what can you do? Become Antifragile To The Punches Whenever we search on Google, we at least glance at the ads before either scrolling past them or clicking on the one that catches our fancy.Now, a fantastic copy might get you the click - cos we’re all wary of deceptively tantalizing copy; what will get you the click and the best chance of conversion is —Brand Recall.If I recognize the AD that shows up on my intent search, there’s a 99% chance I’ll click on their AD instead of anything else.That’s the power of Brand Recall.So, if you’re getting sandwiched by a competitor outspending you on your target keywords, build better brand recall and see your results convert while their ADs remain unclicked! Command Brand Recall Instead of Praying For It HackerNoon has helped 4,000+ companies command Brand Recall in myriad ways, and here’s our boldest campaign offering yet: Site Takeover! The Premise is Simple: Enable our partner brands to capture everlasting Brand Recall without fearing getting sandwiched. At $25k a week, you get: A logo AD atop our homepage
A big-ass banner AD next to our top stories On a CPM basis, it’ll come out to $10,…and if you book a meeting here to confirm a purchase, you get a $5k flat discount per week you buy. If you’re getting sandwiched, that’s your silver bullet, but if you broke the sandwich another way, let us know in the comments. You probably like sandwiches as a snack, but what about being sandwiched by your competitors on Google Search? Elastic Email (a Mailchimp competitor) has been advertising on Google for months. Google Ads constitute ~40% of their ad spend, so they must get fantastic results there, right? WRONG!! If you Google Elastic Email right now, here’s what you’ll see: They. Got. Sandwiched! They. Got. Sandwiched! If you went ‘oooooof’, then you’re a marketer at heart and should read on. Google Likes Your Money - Not You. When Google opened up the bidding mechanism for ADs, the first casualty was Ethics. Until then, Marketers refrained from encroaching upon competitors’ branding, and any harmless banter was considered' a bold move’—remember the Cola Wars? But today, we’re all cutthroats over that $5 CPM. Thanks, Google! Sandwiches like the one above are the norm, not the exception, while Google laughs its way to the bank. So, what can you do? Become Antifragile To The Punches Whenever we search on Google, we at least glance at the ads before either scrolling past them or clicking on the one that catches our fancy. Now, a fantastic copy might get you the click - cos we’re all wary of deceptively tantalizing copy; what will get you the click and the best chance of conversion is — Brand Recall. If I recognize the AD that shows up on my intent search, there’s a 99% chance I’ll click on their AD instead of anything else. That’s the power of Brand Recall. So, if you’re getting sandwiched by a competitor outspending you on your target keywords, build better brand recall and see your results convert while their ADs remain unclicked! Command Brand Recall Instead of Praying For It HackerNoon has helped 4,000+ companies command Brand Recall in myriad ways, and here’s our boldest campaign offering yet: Site Takeover! The Premise is Simple: Enable our partner brands to capture everlasting Brand Recall without fearing getting sandwiched. At $25k a week, you get: A logo AD atop our homepage A big-ass banner AD next to our top stories A logo AD atop our homepage A big-ass banner AD next to our top stories On a CPM basis, it’ll come out to $10,…and if you book a meeting here to confirm a purchase, you get a $5k flat discount per week you buy. if you book a meeting here If you’re getting sandwiched, that’s your silver bullet, but if you broke the sandwich another way, let us know in the comments.

Stories, News & Mentions about NYRIAD

Nyriad’s Stories on HackerNoon