Powered by The Financial Clinic, Change Machine is the revolutionary online platform that significantly improves service to America’s working poor

Change Machine

February 3, 2018 Javascript https://blogs.msdn.microsoft.com/typescript/2018/01/31/announcing-typescript-2-7/ https://m.signalvnoise.com/stimulus-1-0-a-modest-javascript-framework-for-the-html-you-already-have-f04307009130 https://dev-blog.apollodata.com/improved-apollo-client-devtools-eb43258f103f https://medium.com/javascript-scene/graphql-at-the-rest-aurant-f4091054e82a https://medium.com/styled-components/v3-1-0-such-perf-wow-many-streams-c45c434dbd03 https://medium.com/javascript-scene/top-javascript-libraries-tech-to-learn-in-2018-c38028e028e6 https://medium.com/@dalaidunc/10-things-i-love-about-vue-505886ddaff2 https://blog.logrocket.com/the-increasing-nature-of-frontend-complexity-b73c784c09ae https://www.digitalauthority.me/insights/step-step-guide-driving-android-app-downloads-29-01-2018 https://www.cypress.io/ https://medium.com/seek-blog/sketching-in-the-browser-33a7b7aa0526 https://codeburst.io/writing-javascript-with-map-reduce-980602ff2f2f https://medium.freecodecamp.org/how-i-reverse-engineered-a-chrome-extension-to-write-my-own-flask-app-d77f36421106 https://medium.freecodecamp.org/rock-solid-react-js-foundations-a-beginners-guide-c45c93f5a923 https://medium.freecodecamp.org/why-would-you-not-use-typescript-67d0baa3eaca https://hackernoon.com/progressive-web-apps-the-next-step-in-web-app-development-372235bf9a99 https://hackernoon.com/js-wtf-with-math-79da9a941ec1 https://medium.freecodecamp.org/conditional-rendering-in-react-using-ternaries-and-logical-and-7807f53b6935 DevOps https://medium.com/meus-pedidos/spot-instances-in-production-is-it-worth-fcb3e5951973 https://medium.com/google-cloud/kubernetes-101-pods-nodes-containers-and-clusters-c1509e409e16 Cloud & Serverless https://siliconangle.com/blog/2018/01/30/microsoft-launches-azure-event-grid-managing-serverless-computing-functions/ https://medium.com/@PaulDJohnston/container-people-lets-talk-about-serverless-e6ecb5c437cf http://www.rapyder.com/amazon-s3-top-10-things-careful/ https://techcrunch.com/2018/01/31/microsofts-azure-revenue-nearly-doubled-year-over-year-in-its-second-quarter/ https://hackernoon.com/serverless-observability-part-1-new-challenges-to-old-practices-95de1b94d379 https://hackernoon.com/informative-alerts-with-serverless-slack-unfurling-acb6457e3a59 AI https://medium.com/@andrewng/announcing-the-ai-fund-building-transformative-ai-companies-55c008663072 https://www.ibm.com/blogs/research/2018/02/deep-learning-models/ https://towardsdatascience.com/what-used-to-cost-millions-in-2017-can-now-be-solved-with-machine-learning-for-499-month-10b5b4bb1173 https://medium.com/@aldamiz/how-we-grew-from-0-to-4-million-women-on-our-fashion-app-with-a-vertical-machine-learning-approach-f8b7fc0a89d7 https://blog.gyana.co.uk/retail-evolution-gyana-to-define-the-future-of-brick-and-mortar-b877fbe6bbe8 https://machineleraning.blogspot.com/2018/02/top-artificial-intelligence-startups-of.html https://www.sporttechie.com/nvidia-deep-learning-video-analytics-automate-future-player-scouting/ https://www.linkedin.com/pulse/how-become-machine-learning-engineer-path-andrew-nikishaev/ https://towardsdatascience.com/dealing-with-imbalanced-classes-in-machine-learning-d43d6fa19d2 https://medium.com/@dpatil/a-code-of-ethics-for-data-science-cda27d1fac1 https://towardsdatascience.com/only-numpy-implementing-gan-general-adversarial-networks-and-adam-optimizer-using-numpy-with-2a7e4e032021 https://medium.com/modeldepot/transfer-learning-will-radically-change-machine-learning-for-engineers-78732b2bb415 Big Data https://engineering.salesforce.com/how-apache-kafka-inspired-our-platform-events-architecture-2f351fe4cf63 https://wearecitizensadvice.org.uk/markets-dont-work-like-they-used-to-and-people-are-starting-to-notice-af00ed38014d https://towardsdatascience.com/python-basics-for-data-science-6a6c987f2755 Blockchain https://blog.bigchaindb.com/2018-the-year-for-data-driven-blockchains-1b8ce0b5f9 https://medium.com/@iotasuppoter/iota-real-world-use-cases-are-coming-ab1d8240cf09 https://hackernoon.com/build-a-cryptocurrency-price-tracker-in-5-minutes-d66c3d37ad71 https://medium.com/friendz/how-to-register-to-friendz-ico-8d02fe1975bc https://media.consensys.net/state-of-decentralized-exchanges-2018-276dad340c79 https://medium.com/@ryandonz/cryptocurrency-the-secret-to-profiting-from-the-most-volatile-market-in-the-world-via-asymmetric-3fd50e56405 https://hackernoon.com/a-crash-course-in-blockchain-and-crypto-6b6794f41814 https://medium.com/cryptomuse/the-unmasking-of-bitcoin-creator-satoshi-nakamoto-the-fallout-86102a3c60eb Personal Growth https://medium.com/@alltopstartups/how-to-teach-yourself-to-read-an-entire-book-in-a-single-day-485e35e28126 https://medium.com/swlh/i-sucked-at-writing-so-i-spent-a-month-writing-for-4-hours-a-day-dba37aa85585 https://medium.com/the-mission/how-to-discipline-your-mind-for-greater-success-e44d15a2a0ee https://medium.com/the-mission/today-is-the-first-day-of-your-life-a0c9e0b91d05 https://medium.com/swlh/how-to-disrupt-the-daily-grind-slipstreaming-your-dream-be0fae6fc04d https://medium.com/@dariusforoux/15-limiting-beliefs-that-are-holding-you-back-from-reaching-your-full-potential-adc795a4049 https://medium.com/swlh/this-expert-approach-will-prepare-you-for-any-interview-d8b62e3be03d https://medium.com/the-mission/when-you-love-yourself-first-life-will-take-care-of-the-rest-7e8c7e9b3097 https://medium.com/@RyanHoliday/the-moments-no-one-wants-to-experience-but-everyone-needs-to-succeed-c284273b3d59 Startup https://medium.com/@jc_70515/growth-hack-your-way-to-the-top-of-product-hunt-2f09e925cc06 https://medium.com/@zebulgar/how-to-raise-money-before-launch-a3544ef4dba6 https://medium.com/swlh/steps-to-validate-your-startup-idea-before-you-spend-any-even-1-43723781afd3 Misc https://medium.com/@jamesflight/critical-thinking-in-software-development-the-word-should-and-why-you-shouldn-t-listen-to-563090144331 https://www.simform.com/ecosystem-serverless-technologies/ https://medium.com/berkman-klein-center/the-american-museum-of-exploding-cars-and-toys-that-kill-you-5123f35cb271 https://towardsdatascience.com/16-useful-advices-for-aspiring-data-scientists-6da9afa8c72c

Week 5: Reading list & news to keep up-to-date with Javascript, Python, DevOps, Cloud, Serverless…

By , Founder and CEO at Washlava. on . Todd Belveal Originally published Quora We all want our startups to be the next Uber, the next Airbnb. Everyone loves a good underdog story. Whether it’s a 15-seed running deep into your March Madness bracket, or an NFL quarterback who’s working at a grocery store to keep his dream alive, we’re all eager to see the little guy win every now and then. And some of us get a similar thrill when a startup muscles its way into an industry, going up against established players and coming out on top. Everyone is looking at industries with established companies and wondering how they can topple them. The question is, is it possible? Is this industry ripe for disruption? I’ve founded two startups in two very different industries — rental cars and laundry — and I can tell you, there are some signs. There are things that will tip you off that an industry could use some new blood. 1. Power Is Consolidated One of the important signs that an industry could be disrupted is imbalance, or dominance by one side of the economic equation. Oligopolies, where a few companies have consolidated vast amounts of the market share either on the supply or demand side, are often good candidates. Rental car companies are a great example. Nine brands control 80% of the market. Not too bad, right? But there are actually just three companies that own those nine brands. Two of those company headquarters are actually in the same little town in New Jersey. It’s incredibly consolidated. And that makes it seem difficult to disrupt because the barriers to entry look high. But that sort of entrenched, oligopolistic behavior (despite competition) generally leads to a greater incentive to preserve the past rather than innovate for the future. So, when things do change, those huge companies can stumble. They’re big, slow, and often too complex to adapt to new technologies. That’s where startups shine. Startups have the ability to work fast, to develop quickly. Speed is their greatest asset, other than their technology. You may be a small fry, but you’ll likely find opportunities when you enter an industry with these dynamics. A lot of the large companies in these industries have become more self-aware, and you may even find an exit among these important players. 2. Consumers Are Using Outdated Technology Massive companies often have incredibly complex systems, and that makes it difficult for them to implement new technology quickly. Creating a smartphone app seems pretty simple, until you realize you have to deal with your complicated backend system and business model. A mobile platform requires simplicity, not complexity. Industries where the major players are stuck using old technology are usually ripe for disruption. A good example of this is the laundry industry. I know, disrupting laundromats doesn’t sound flashy, but that’s exactly the point. Laundry isn’t flashy. It’s old school. Whether you’re doing it in your apartment building, or you’re heading to the laundromat, there’s a good chance you’re still using coins. At best, you might have a rechargeable card you can swipe. My current startup, Washlava, brings together the machines, the technology, and the users. Consumers can use the app to check which washers and dryers are open, reserve them, and then pay with their phone at the laundromat. No more hoarding quarters. No more trying to force your wrinkled dollar into a change machine. It isn’t the craziest idea anyone’s ever had, but it feels like it because of how antiquated the current system is. Old technology is often used. Not because it’s the best way to do things for consumers, but because it’s the best way to do things for a few entrenched business interests. That’s an opportunity for someone to give consumers what they really want. 3. Business Practices Aren’t Changing — Despite Negative Consumer Sentiment When developing the original concept of Silvercar, what struck me was just how universally panned the car rental industry was. I was obviously inspired by my frustrations, but I was not alone. Car Rental has a net promoter score of 23. (Just to give you a benchmark if you’re unfamiliar with the NPS, Apple has a net promoter above 80). Basically less than 1 in 4 rental car customers in the US would recommend these brands to a friend. At Silvercar, we were working to create a simplified experience for renting a car. Something that might not be for everyone. But for the right person — say, a business traveler — it would feel like a godsend. Despite the sentiments reflected above, car rental companies did very little to improve the overall experience — either because they didn’t want to, or simply couldn’t due to the complexity of their systems and operating models. After Silvercar was launched, National rolled out their “choose any car on the lot” campaign. It was a smart move because choosing a car was part of our value proposition, but it was also an easy move. It wasn’t exactly a game changer, because you’re still choosing from the unexciting selection of cars on the lot. But that type of fealty to the status quo is exactly why companies are sluggish when they have to respond to changes in the market. It’s exactly why startups can flourish in those markets. 4. The Research Backs You Up I always recommend that people spend a lot of time, and even a lot of money, doing research on the industry and the idea. Why? Because research breeds confidence. Confidence in your idea and confidence in your business plan. Spend what you need to make sure there’s really an opportunity. Figure out who the competitors are and what they have. You need to know what you’re getting into from a competitive standpoint, and you need to know that your company could truly contribute to the industry. The research is what will confirm all the other points I’ve talked about. Or it may tell you that breaking in will be tougher than you thought. Whatever the case, put in the time to research what you’re getting into. It’ll give you the confidence you need to take your shot at disrupting an entire industry. By , Founder and CEO at Washlava. on . Todd Belveal Originally published Quora For more trending tech answers from , visit . Quora HackerNoon.com/quora

Is This Industry Ripe for Disruption?

This guide sheds light on the workings of Virtual Private Networks (VPNs). While commercial providers often overpromise, it's essential to grasp the fundamental mechanisms behind their capabilities. We'll use plain-language explanations and simplified Python code samples to illustrate core principles. These simulations showcase concepts on a smaller scale—true VPNs implement IP masking, encryption, and tunneling at deeper system levels for robust real-world protection. Important Note: Core VPN Fundamentals IP Address Masking What IP masking is Terminology: IP Obfuscation, Proxy Simplified Python Simulation Data Encryption Encryption as a data scrambler; emphasis on confidentiality. Terminology: Encryption protocols (AES, etc.) Simplified Python Simulation Secure Connections (Tunneling) The "private pipe" analogy for establishing secure tunnels. Terminology: Tunneling protocols (OpenVPN, Wireguard, etc.) Simplified Python Simulation What VPNs Can (and Cannot) Do Enhanced Online Privacy Secure Data Transmission Bypassing Geo-restrictions (no complete anonymity, malware protection, etc.)' Limitations of VPNs 1. IP Address Masking Mechanism Explanation Picture your internet traffic as letters you want to mail. Your IP address is your return address on these letters—and your ISP (Internet Service Provider) is like your local post office. They handle and see all your internet traffic coming and going. VPNs act like a special postal forwarding service. When you connect to a VPN, your letters get rerouted through their hub - the VPN server. Now, instead of your home address, those letters bear the address of the VPN's postal hub. That's how websites see the VPN server's IP instead of yours. Think of "masking" as putting on a disguise. IP masking hides your real IP address, making it harder to link your online activity back to you directly. IP Obfuscation/Masking: A proxy is like a temporary middleman. For this example, a proxy server will briefly make it look like your internet traffic is coming from location instead of yours. Proxy: its Simplified Python 'Simulation' Example While Python can't truly change your machine's IP, we can mimic the idea with a proxy: import requests Simulating IP Masking by Using a Proxy proxy = {
  "http": "http://your-proxy-address:port",
  "https": "http://your-proxy-address:port",
} url = "http://ipinfo.io/ip"
response = requests.get(url, proxies=proxy)
print("Masked IP Address:", response.text.strip()) 2. Data Encryption Mechanism Explanation Imagine your internet data as letters containing sensitive information. Data encryption is like scrambling those letters into a secret code using complex math puzzles. To make sense of the scrambled message, you need a special key. VPNs use tough encryption methods to turn readable data (like login details or bank info) into this messy code. If anyone (like a snoop on public Wi-Fi) tries to peek at your traffic, all they find is gibberish. Only you and the VPN server have matching keys to 'lock' (encrypt) and 'unlock' (decrypt) the data. Shared Secrets: Simplified Python 'Simulation' Example Python won't mimic true VPN encryption, but let's see the principle: import socket
import ssl Create a Secure Socket Using SSL context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH)
secure_socket = context.wrap_socket(socket.socket(socket.AF_INET), server_hostname="www.example.com") secure_socket.connect(("www.example.com", 443))
secure_socket.sendall("GET / HTTP/1.1\ Host: www.example.com\ ".encode()) Receive Data Securely response = secure_socket.recv(4096)
print(response.decode()) secure_socket.close() Real VPNs use encryption (like AES) that's far harder to crack than our example. Strong Encryption: Terminology: Encryption Protocols (AES, etc.) AES (Advanced Encryption Standard): Commonly used in VPNs, renowned for its strong security. Key Points Encryption makes it extremely difficult for outsiders to understand intercepted data. This ensures sensitive information stays safe in transit, especially on public networks. 3. Secure Connections (Tunneling) Mechanism Explanation: VPN tunneling involves encapsulating and encrypting each data packet sent over the internet between your device and the VPN server. This secure "tunnel" prevents outsiders from intercepting, viewing, or modifying the transmitted data. Tunneling protocols, like OpenVPN, establish this secure connection by using encryption and other security measures. Python Sample Code: Simulating a VPN tunnel in Python is complex and beyond the scope of simple scripts due to the networking and system-level operations involved. However, you can use Python to create a basic secure connection using SSL with the library. socket import socket
import ssl Create a Secure Socket Using SSL context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH)
secure_socket = context.wrap_socket(socket.socket(socket.AF_INET), server_hostname="www.example.com") secure_socket.connect(("www.example.com", 443))
secure_socket.sendall("GET / HTTP/1.1\ Host: www.example.com\ ".encode()) Receive Data Securely response = secure_socket.recv(4096)
print(response.decode()) secure_socket.close() Key Points Tunneling defends against snooping, data tampering, and other risks associated with untrusted networks. Terminology: Tunneling Protocols (OpenVPN, Wireguard, etc.) Popular, flexible, and secure. OpenVPN: Newer, highly streamlined protocol gaining favor for speed and security. WireGuard: Core network-level protocol often used to secure VPN tunnels. IPsec: What VPNs Can (and Cannot) Do VPNs Enhance Your Online Experience IP masking and data encryption make it substantially harder for advertisers, your Internet Service Provider (ISP), and potential snoopers to connect your online actions back to you. Enhanced Online Privacy: Vital on public Wi-Fi! A VPN safeguards your logins, financial details, and other sensitive information that could be easily intercepted on an unsecured network. Secure Data Transmission: If a streaming service, website, or online content is region-locked, connecting to a VPN server in the appropriate location often helps you access as if you were there. Bypassing Geo-restrictions: What VPNs Cannot Do Although they make you much harder to pinpoint, factors like browser fingerprinting, cookies, or your online behavior patterns can still leave traces. Guarantee Complete Anonymity: A VPN secures your connection, but it's not a cure-all. Downloading infected files or falling for phishing links remains dangerous, even when using a VPN. Protect Against All Forms of Malware and Phishing Attacks: A VPN is one powerful layer of protection but doesn't substitute common-sense online safety. Stay vigilant about suspicious websites, links, and emails. Always be cautious, even when connected to a VPN. Eliminate the Need for Safe Browsing Practices: Important to Remember When selecting a VPN provider, carefully consider their reputation, features, logging policies, and commitment to user security. Not all VPN services are equal in the protection they provide. It's wise to thoroughly research options before investing your trust.

How VPNs Work: A Simplified Breakdown for Non-Techies

There is a general debate going on how ethical or unethical the use of AI is, however not many people are talking about the challenges in adoption of AI by Small and Medium-sized enterprises. So, before we go one pondering about how people will lose their jobs due to AI , or before we actually start looking for new careers without actually knowing what  AI is about, let me take you through a few challenges we are facing in the implementation of Machine learning and Deep learning programs and apps developed on AI platforms, in the real world especially by the majority of businesses around the globe. Widespread AI Phobia and insecurities caused by Sci-fi books and films AI phobia is not a new kind of fear, it is a fear which we have been living with all our lives due to the irrational works of fiction writers and movies. This fear has been around long before the technology was even developed if you have watched movies like Terminator, you know exactly what I am talking about. This phobia is so rampant that even great minds like Stephen Hawkings and Elon Musk have been very vocal about their irrational fear of AI. Musk has been quoted several times that AI is more dangerous than nukes. So the question here is…Should we give machines the power to learn? Actually yes, according to experts in Machine learning A.I, this is kind of fear is very natural and we as humans have always resisted change and machines can only learn what you have programmed them to learn. Email Security Expert, Talha Obaid who has worked extensively on developing Machine learning A.I and has patents of his own, believes that AI can never actually replace humans. “Think about it this way, although self-driving cars have already been invented, we prefer using the GPS instead,” Obaid said, before adding that its human nature, that human species will never be comfortable letting go of the reigns. Obaid also stated that AI cannot ever be smarter than the human mind and can only learn to do the things we program them to learn, they might have glitches but as long as the person creating them does not have an unethical reason to create them, the technology itself is not unethical. “An AI robot may be able to brew me a cup of tea or fix me lunch, but that does not mean it will replace my wife or mom in the kitchen. It will only be used to help them out with the chores and not wipe them out of their kitchens,” he said. The financially back-breaking and lengthy development process AI with Machine learning capabilities cannot be created in a day or two or even in a week or two, it requires a lot of time and effort. It also requires companies to hire experts in AI technology, to work with their team. This can turn out to be exceptionally expensive as many developers find it difficult to learn Machine learning programming on AI and in turn, there is a void in the number of experts who are proficient with machine learning and so hiring them can be quite expensive. Although we may argue that this is a one-time investment, many startups and Small and medium-sized enterprises do not have surplus cash for innovations, no matter how successful they are. Similarly, developing a Machine learning projects on AI platform, app or device is a lengthy process. According to Obaid, you need two teams involved one team will have the expertise on the subject that the AI is being developed for and the other will be proficient in AI. “ Most companies already have domain experts who can provide the AI team with the required data for them to build the AI technology that the company requires,” he said. He further explained that this can be a tedious and rather long process because the exchanges between the two teams are generally never-ending. “The domain experts know what the final product should look like and they need to provide the AI team with the information, one slip in providing the right information could be a recipe for disaster. So there is always a lot of exchange between the two teams, which is very time consuming and the result is not always as expected. In fact, according to a recent report by , only 5 percent of businesses worldwide have adopted implemented AI technology successfully. According to him, things could get much simpler if the domain experts were trained in house on Machine learning AI. “I am saying this from my own experience, if the domain team is trained in AI than it saves the company a lot of time and money because it would be just one team handling the whole development process,” he explained, before adding that the lines between domain and AI experts need to be blurred for successful implementation of the technology for business purpose Mc Kinsey Return of Investments uncertain Many small and medium-sized enterprises do not have enough cash flow to develop and use their own AI technology and this is why digitization of this nature at this point is mostly prevalent in MNCs and other corporate players with big money. Also apart from the cost-based scenario, many businesses still lack faith in the difference that AI can make in their business. “Business is all about making money and return on investments are extremely important, so many businesses are just not ready to take the plunge,” said Israeli Pre-Seed investor and Quality Assurance Expert Amos Raber. “Most businesses according to him fail to use AI technology appropriately because they tend to jump into the bandwagon and not align their businesses with the technology they require to earn them profits”, he added. So, although AI one day, maybe used in every sector, successful implementation of this technology in the real world has become a challenge and though we do have the technology to make our lives simpler, we just haven’t been able to provide it to the people who need it.

Challenges in successful implementation of Machine Learning AI in SMEs

Understanding intent and form to develop a palpable interpretation of what’s rational and irrational. From almost instinctual guidance, consider what ‘snaps’ your attention, what strangles your gaze. What’s abnormal isn’t sensible. Our sense of attention fixates on the ‘distinguishable’. This beam of correlation ensures that we can function properly in new environments, extrapolate meaning in unfamiliarity and render art when seemingly handicapped. We can fill in the blanks. Such insight allows us to bypass more traditional methods of learning — rote learning. Those that seem smarter can generally complete more with less effort — they can fill in the blanks that the others need spelt out. While others are reactionary to their world, there’s some that can march into the unknown and return as conquerers. Clearly rote learning isn’t the only way a machine can learn. I’m near the final stages of a personalised food-ordering app that’s taken the better part of 2.5 years to develop. Now as I begin developing peripheral services like customer-service and onboarding-processes, I’m trying to automate as much as possible to minimise overheads. Given the nature of the app, it’s likely there’ll be high engagement between users and customer-services so any automation needs to be functional and at a minimum, on par, with a human operator. I need to build a chatbot. And it needs to work. In short, I want it to be revolutionary. really I’m familiar with the general ways of teaching a chatbot: broadly leveraging NLP/NLU and parsing sentiment, sentence/intents, dialogue etc, but I’ve always found that most ML based techniques are just rote-learning on steroids and I don’t like that. Time and time again, programmers are told how toddlers can outsmart state-of-the-art ML algorithms and that the only way to outpace this crawler is to nuke their computer with billions of data-points. Upgrade to a new graphics card. Buy a new Intel multi-core processor. Train through a VM on AWS. The list is endless… Yet the suggestions are still dumb. We should be focusing on understanding the fundamental rules this data obliges in an open-context rather than dissecting the individual pen-strokes of the number ‘9’ in MNIST. Sure low-level insight is needed to interpret stimuli, but the scope of ML needs to grow from simple, closed-based arenas to the battlefronts of the real-world where information is randomly distributed and spontaneous. To make the first step into this gunfire, we need to change how machines ‘think’. Sensical Filtering A network of competing logic-modules that test their understanding through emulation. Over the coming weeks I’ll be implementing this idea and once completed, I’ll return with notes. For now, here’s my general goals: The goal is to develop a system that can understand the intent of a user’s message without being trained on a corpus spanning millions of examples. The system should be able to understand associations of speech and how they precipitate actions. That is how for example, escalating rhetoric can violently precipitate confrontation whereas a deescalation can reverse tension. Interpretations construct a growing sense of understanding that’s relevant to the temporal and spatial properties of the conversation. The more we hear, the more we know. There’s a hierarchy of information, organised with respect to scope. Information not only contributes in clarifying this hierarchy but allows for better searching policies as the scope of conversations can vary, meaning the relevance of information is dynamic and related to the intent-at-hand. As for the implementation: Parse a text, such as a book, speech, screenplay and associate how dialogue leads to actions. Exploit the natural structure of the text to represent sequences. Higher-quality writing will provide greater boundaries between semantics. Assess the formation and degradation of sequences across the text, with each sequence escalating or deescalating the intents that have been pre-built. Intents will then gradually become attributed to particular phrases. Using TF-IDF and cosine-similarity, the respective distinction between phrases/words can be attributed to ideas. Using this hierarchy of scores from TF-IDF, a tree of intents can be developed. Using this tree-like structure, we can employ a Monte-Carlo searching algorithm during conversations with users to understand and predict what users to discuss. want Training should involve emulating the derived insights and seeing how they diverge in unseen texts. That is, being able to ‘write’ their own text and then finding texts that are similar to this written piece. This would be analogous to daydreaming/imagining etc. It’s still early days so it’s unknown how this will fare, but nonetheless I’ll be sure to write up the results once I’ve implemented the code. Thanks for reading!

Marching to Programmable-Intelligence

As many still worry about how robots are rapidly going to replace humans for repetitive and monotonous activities, Wall Street’s key players are already applying Artificial Intelligence for operations that until now would have required the most skilled financial brains available on the market. Just as the name Machine Learning suggests, robots are now able not only to obey to our commands but also to learn and deeply analyze what they are being trained for, until they master the matter and they manage to self-improve, without the need of any kind of human intervention. Consequently, the main idea is not to build machines that one day will take over us, but rather provide humans with a valuable assistant for all those procedures for which we could really take advantage of some extra help. Although we might not like to admit it, we cannot strive for progress by ourselves, as the resources of the human nature are limited- just think about our time, our ability to stay focused and our processing speed. A good turnaround for it is to collect our energies and organize them to build a set of algorithms that can make up for all our flaws and the characteristics that humans are missing. Be careful though, as the Machine Learning topic is nowadays often misused and brought out in matters in which it might be not relevant or related at all. Truth is that, as Artificial Intelligence researcher Eliezer Yudkowsky states, “By far the greatest danger of Artificial Intelligence is that people conclude too early that they understand it.” Artificial Intelligence can indeed be applied in the most diverse fields, but we need to clearly have in mind the matter, and see how each industry can benefit differently from this technology. Quantitative disciplines, more specifically, are the ones that could benefit the most from Artificial Intelligence advancements. Due to the great amount of big data that its industry has, the financial sector is particularly fertile for the implementation of Artificial Intelligence and Machine Learning. Artificial Intelligence and finance, in fact, represent a good match for many different reasons. Retail banks are already looking into AI for the realization of more advanced chatbots, robo advisors, and compliance tools.After having conducted a survey among 86 banks, UBS came to draft a prediction according to which, in the next couple of years, Artificial Intelligence could increase banks’ revenues by 3.4% and lower their costs by 3.9%. Following this path, in the autumn of 2016 UBS SmartWealth was launched, with the purpose of providing real-time financial advice and suggesting personalize investment strategies. On the same year, Capital One and Bank of America came up with similar products. The first one launched a particular skill, aimed at helping customers manage their credit card and bank accounts through Alexa, Amazon’s virtual assistant, that, just like Apple’s Siri and Window’s Cortana, replies to — almost — every question we have. BoA, on its side, introduced Erica as its clients’ new financial assistant, easily reachable through the bank’s mobile application, and able to help with all the possible daily tasks, such as balance checking and money transferring. Another financial field in which AI can play a key role in fraud detection, that can be accomplished with the analysis of consumers’ behavior, with a heavy focus on suspicious transactions. The relationship between AI and finance, however, goes far beyond the assistance in daily operations and the automation of certain tasks. As we anticipated before, the potential of its technology could arrive at the point of overcoming human abilities and accomplishing incredibly difficult tasks. For the amount of data available and for the complexity of the matter, trading is the financial branch that could be benefitting the most from the intervention of AI algorithms. Just think about it: the stock market opens when the day starts in Sydney and closes when it ends in New York City. Basically, there are transactions happening 24 hours per day, 7 days a week, excluding the few hours during which the weekend is happening all over the World. Every day registers thousands of millions of transactions, and this makes the market extremely active and subject to sudden changes. Decisions have to be taken rapidly and even the most experienced traders often face difficulties and make fatal errors in their predictions of the trends and in their forecasts. Human beings might find hard to accept to change a strategy they have pondered so much before. Machines, on the other hand, do not have that sentimental problem that makes them feel like their “married” to their original idea and they can’t change it. Machines’ algorithms are programmed with the purpose of making them able to rapidly decide when to buy, sell or hold a stock, with a much better accuracy compared to human beings. A big limitation to the implementation of AI algorithms for trading is represented by the lack of transparency that this system involves. Machine Learning developers might not only deny sharing their technical secrets but could also find it hard to explain them, as, the more developed the algorithms are, the harder it is to make them understandable for someone that doesn’t have that much of an AI background. This could represent a big limitation for the expansion of AI in trading, as, if the process is not fully understood, not everyone would agree to invest a relevant sum of money and give the entire control of it to a robot. For this reason, you need to have a trusted partner on your side, that gives you all the information you need in a transparent way and that makes you feel comfortable while making your investments. Most importantly, you need to have a user-friendly and clear platform, that explains the algorithms in a way that you can easily understand the conclusion that they came at and the process that guided them to it. might be just what you need, follow us on our website to find out how our system will work and participate to our Q&A sections to clarify all your doubts. AITrading Any questions or concerns? Join our Telegram community and Telegram channel! Wanna get more trading signals? Join our channel on Telegram! Follow our social media resources. AITrading links Website Roadmap Whitepaper Medium Blog Telegram community Telegram channel Twitter Facebook Insta Bitcointalk ANN Steemit Subreddit

AI and ML influence on the worldwide finance and trading environment

Securing your network connections using OpenVPN

The of my Amazon Web Services application was supposed to be It’s two independent Lambda functions with a Redis cache between them for persistence. By purchasing Lambda compute power by the millisecond, I thought my app would be stupid cheap, possibly even free! I hadn’t even launched into production when I got my first bill for my partial month of development: $26.97 and if that wasn’t bad enough, my forecast for the following month was supposed to be $40.78 and nobody had even used my app! architecture simple . The shock of this discovery taught me a valuable lesson about creating hobby projects on : . In the rest of this essay I’ll walk through how I identified the costlier parts of my app, and what I did to trim costs. AWS Beware of over-provisioning When I received an unexpectedly nasty bill from Amazon, my first reaction was to not reopen the email while cursing. “There has to be a mistake…” I say in complete denial. The first step should be to use the billing console on the AWS website to identify where the costs are actually coming from. When I went to mine, I found the following scary picture: Go to the billing console! $40 for an app I haven’t even launched, yikes! From there, I noticed two immediate cost centers: EC2 and Elasticache. That was weird, I thought. I’m not even using any EC2 instances, I’m only using Lambda, and my Redis node in Elasticache only stores 5 strings. How could that be costing me almost $41 a month?! From the main page I jumped to the dashboard where I could clearly see what was costing me the money: Aha! My two targets. I could see that it was clearly two problematic services, my NAT Gateway and my Redis node. With cost reduction in mind, I knew I had my work cut out for me. My first job was to replace the costly NAT gateway that allowed my Lambda to connect to external networks. Amazon provisions these gateways to deal with bursts of up to 10 Gbps. That was way over provisioned for my simple needs. So I found an interesting : setting up a free t2.micro with a NAT AMI. Luckily Amazon provides a preconfigured image exactly for this purpose ( ) I spun up an instance, and put that it into my VPC in a subnet with an internet gateway. Rolling my own NAT Gateway: solution amzn-ami-vpc-nat-hvm-2016.03.0.x86_64-ebs . The new instance goes into the subnet with the internet gateway as a default route. Amazingly this worked right out of the box! I’m thankful that Amazon provides a machine image to duplicate their NAT gateway functionality, but I find it unfortunate that it is so difficult to set up a Lambda function to make outbound requests. If you are having trouble with Lambdas and VPCs, I recommend checking out this AWS , and Vincent de Lagabbe’s very helpful StackOverflow . article answer Cost of running new t2.micro instance: $0 : The second cost center was my Elasticache Redis instance. The reason it cost me $6 was that it was running on a t2.small instance. Considering my simple use case, I just needed to move it to a t2.micro instance. Just change the machine type in the Elasticache console right? . Unknown to me, I had used the default config for Redis on a t2.small Elasticache cluster with two nodes. A t2.micro only supports one node. Running Redis on the cheap Wrong This wasn’t documented anywhere and to add to the frustration the AWS console would allow me to change the cluster machine type, but just wouldn’t show the T2.micro anywhere in the list. It took me running commands directly from the AWS CLI, to discover that in fact, I could not change my Node type to micro: aws elasticache list-allowed-node-type-modifications — cache-cluster-id roadbotomicro The solution was simply standing up an entirely new Elasticache instance with on a t2.micro instance**.** one node Cost of Redis on t2.micro: $o : AWS is a fantastic platform for hobbyists. It is also a financial behemoth that makes gobs of money by over provisioning its resources for big companies that forget about how much these little expense add up. However, as a hobbyist you need to be extra careful about your own budget. Much the same way we have learned to be cautious with installers or download buttons, we need to be careful about the default configuration for AWS products. Moral of the story When you avoid the shiny defaults…

CHANGE MACHINE

change-machine.org

-- employees

Since n.d.

COMPANY RANKING

CHANGE MACHINE

EVERGREEN INDEX #1979

Representing this company? Click here to claim and customize this page!

Change Machine WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

CHANGE MACHINE

change-machine.org

-- employees

Since n.d.

COMPANY RANKING

CHANGE MACHINE

EVERGREEN INDEX #1979

RELATED COMPANIESRANK

STORIES ABOUT CHANGE MACHINE

Representing this company? Click here to claim and customize this page!

Change Machine WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!