Fulfilling the administration’s vision of a digitally connected “Bagong Pilipinas,” President Ferdinand R. Marcos Jr. launched on Thursday the Philippine Domestic Submarine Cable Network (PDSCN)…

Presidential Communications Office

MANILA – The Philippine government is racing to repatriate 14 out of the 15 Filipinos still in Gaza amid reports of Israel’s planned military operation in Rafah. Philippine Ambassador to Jordan Wilfredo Santos said Thursday the Philippine embassies in Jordan, Egypt, and Israel are closely…

PH to repatriate 14 in Gaza; remaining Filipino nun ‘unreachable’ | Philippine News Agency

The tensions between China and Taiwan continue unabated. The latest such incident off the coast of Kinmen resulted in the deaths of Chinese fishermen and brought to the fore official statements from both countries stating they took the right action. Taiwan apart, China has expelled a Filipino ...

Blood Spills As China-Taiwan Tensions Escalate; PLA Expels Filipino Vessel From Contested Waters

'She is a pioneer in shattering gender stereotypes, venturing into tattooing when it was just a man’s exclusive preserve,' President Ferdinand Marcos Jr. says of Apo Whang-Od

Filipino tattoo artist Whang-Od awarded the Presidential Medal of Merit

Fliptionis in the Video Creation Tools Industry

Fliption

A time is coming when your thoughts will be the only activity you need to get things done. Have you ever been so caught up with work, you don't want to move a muscle? You are not willing to use your voice because you feel the ideas will escape once you do. So, what do you do when you suddenly wish your favorite songs were playing at that moment, but you feel too lazy to pick up your phone or call Siri's attention to turn on your playlist? If only your thoughts could make the orders, right? Here, you are thinking, "Siri, play me 'Easy on Me' by Adele" and seconds later the sound fills your workspace. Then you think, "Nope, too low. Siri, increase the volume to 10", and the volume gets adjusted to suit what’s on your mind perfectly. Your brain alone processes commands without any effort from your mouth. I guess that can be seen as a grand breakthrough in the history of neuroscience—or probably not! Let's find out what this means for the world collectively. What Exactly is Neurocomputing? Neurocomputing is a branch of neuroscience that deals with the examination of brain patterns and activities. It is a subfield of cognitive neuroscience that scientifically studies the biological mechanisms and reactions of the brain to different experiences. The term applies to the process of reading, analyzing, and interpreting brain activities by utilizing brain sensors and an imaging framework to examine how our brain reacts in response to different emotions. This is usually through the use of a Brain-Computer Interface Device. The framework, in this case, is called the "triad,” a 3-step process of emotion-valuation, sensory-motor, and meaning-knowledge evaluation. With the introduction of Al, neuroscientists hit a new milestone with brain activity studies. The Rise of Neurocomputing Would you have ever imagined that one day, the content of your thoughts could be read by devices? Science fiction? Sure! Why not? We see a lot of that. But reality? That's almost impossible (or so we thought). Lie detectors were the most mind readers, but they can never detect what your thoughts are word by word. However, technology always seems to awe us as time progresses because the technology we once thought only existed in movies is right here! The use of brainwave-controlled devices to read minds began with the invention of Electro Encephalogram (EEG) in the 1900s. History has it that the German psychiatrist, Hans Berger, created the in neurosurgery. first EEG brain recording These devices, fitted with sensors, were used mostly to monitor and record a patient's brain activity. When placed on your scalp, the sensors pick up electrical signals produced by the brain, which can be interpreted. Around 2008, consumer-focused EEG became commercialized for public use. and —both tech companies—were among the early players in this scene as they released EEG headsets with the promise to create mind-controlled video games. NeuroSky Emotive Furthermore, it began being tested in other areas besides video games, like how it could be used to control your phone or other activities by companies like and . iBrain Muse Relief from Brain-Computer Interfaces Technology Innovations and advanced companies like Meta and Neuralink notched it up with their research into Brain-Computer Interfaces (BCIs) - a technology that can pick thoughts directly from brain neurons and translate them into words. This technology uses AI algorithms to read brain activity and decode emotions into words. Between 2017 and 2019, Meta began funding . As of March 2022, the funded Nature Neuroscience, UCSF researchers achieved brain-to-text decoding with an average error rate as low as 3% when it was tested with vocabularies of up to 300 words—a huge advancement from previous studies. this technology research into At the same time, Elon Musk's Neuralink is already making a good deal of progress with its flexible ultra-thin threads that can be implanted directly into the brain, enabling the carrier to control their gadgets/devices with their thoughts. This is already being tested on paraplegics. The technology has been very much welcomed by paralytic patients who need to gain a good level of control over their lives again. Neuralink has already begun testing the brain threads on paraplegics. Case Study of Neurocomputing Although this technological concept sounds somewhat new, neurocomputing has been in the research and development stage for years, and test-flight gadgets are already being produced. The possibilities of mind reading and brainwave-controlled gadgets are endless. Scientists continue to look into ways to integrate them into our lives, experimenting across various sectors with the possible use of non-invasive EEG devices and BCIs. Many of these experiments are often procedures aimed at helping the physically and mentally disabled to communicate their needs and gain some control over their lives. The world is not new to cases of patients who can not speak due to sicknesses, paralysis, stroke, or just being born that way. So when, for the first time in history, researchers at the GrapheneX-UTS Human-centric Artificial Intelligence Centre, University of Technology Sydney, invented a portable non-invasive device that could translate thoughts into words, it was seen as a breakthrough in neurocomputing. Breakthrough at the GrapheneX-UTS Human-centric Artificial Intelligence Centre: At the time, though, the accuracy was questionable, but with the initial goal achieved, it would only take more years of developing it to increase its accuracy rating. We already find Meta and Neuralink taking up this task with BCIs and electron threads. Evolving neurocomputing also makes smooth communication possible between humans and machines, be it our gadgets, robots, or prosthetic arms. The history of neurocomputing and brain-controlled devices left a remarkable mark in the FIFA 2014 World Cup when, for the first time, the symbolic kick-off event for the tournament was led and completed by a paraplegic man using an advanced assisted mobility technology—a brain-controlled exoskeleton. The FIFA 2014 World Cup Kick-Off Event: This technology, by , a collaborative invention of over 100 scientists from around the world, made use of a set of non-invasive electrodes to read the brain signals from the paraplegic and transmit them to the lightweight exoskeleton triggering movements that helped the man complete the kick-off successfully. The Walk Again Project In the same year of the FIFA World Cup kick-off event, the director of the Center for Neuromodulation, Ohio State University, Dr. Ali Rezai, performed a on 26-year-old quadriplegic Ian Burkhart. Paralyzed Ian Burkhart Regained Mobility: life-changing surgery The surgery involved implanting a tiny 4 x 4 mm microchip into Burkhart's motor cortex, used alongside an electrical sleeve and purpose-made algorithms. This allowed the sensors to trigger his thoughts to control his hands and fingers while bypassing his damaged spinal cord. It was a success. This is not the only successful brain implant in the history of neurocomputing; a man paralyzed from the shoulders down has also been able to regain control of his paralyzed muscles by bypassing the injured spinal cord with the into his motor cortex. implantation of two aspirin-sized 96-channel electrode arrays EEG devices and BCIs are greatly contributing to improved quality of life for disabled people. Other Notable Strides: But we can't say that brain-controlled and mind-reading devices are only for the sick and disabled. These devices could also be used to carry out other tasks/activities for the good of everyone: BBC and This Place teamed up to create a mind-reading headset to enable its user to control the with their thoughts—goodbye to remote controls. BBC iPlayer A team led by an Associate Professor at the KU-KIST Graduate School of Converging Science and Technology, Korea University, . Suk-Won Hwang, developed E-Glasses This prototype eyeglass is made up of flexible electrode sensors located between the ears and eyes of the wearer. First, the electrodes consist of an EEG to monitor the brain's electrical activity and an ElectroOculogram (EOG) to track eye movements and transmit both from the glasses to a device for processing and interpretation. This glass can be used to monitor the mental health of the wearer, help them control games with their eyes, and turn on and off the sunglasses mode when the wearer needs it. Nissan and NASA are working together to build a self-driving robot car technology called that can read, monitor, and measure a driver's brain activity to know if he/she is alert to drive or not. Mind Sense In an exciting twist of things, in 2013, NASA came up with the crazy idea of using . Together with scientists from the University of Essex, they launched a project to control a virtual spaceship with the mind. BCIs to control planetary rovers A competition was held in 2016 by neuroscientists at the University of Florida, a competition that had pilot flying their drones to the finish line by using their minds. Brain Drone Race And the list goes on and on. But like every other technology and thing that exists, there will always be pros and cons to consider. The Ethical Considerations of Neurocomputing Neurocomputing is a technology that is already transforming lives, especially for paraplegics, who, in many cases, have been able to get robotic body parts that can be controlled with their minds or bypass spinal cord injuries to stimulate their bodies into motion through brain neuron implants. People who couldn't talk can now communicate their feelings and what they want/need and interact through mind-reading gadgets. For everyone else abled people, organizations, and companies that wish to apply these gadgets in their daily lives or their various sectors, it is very much achievable. It can bring "ease" into certain activities and, perhaps, create a more quiet world. However, the implantation of invasive BCIs into the brain involves a surgical procedure that is almost too risky to be considered. The brain is a fragile organ; damage to it can lead to mental abnormalities or even brain death. As such, the risk of brain implants makes this technology far-fetched for fully functional humans. Paraplegics, on the other hand, can wager these procedures because of the quest to have a better quality of life. A delicate situation, but if that is what it takes for them not to be stuck in a static world at the mercy of others, it is worth the risky shot. This is why BCIs by companies like Neuralink and Meta, although envisioned to cater to all and sundry, are more appealing to the paralytic section of the world at the moment. Because of these concerns, scientists are looking back into non-invasive methods of mind reading. Last year, researchers at the , University of Technology Sydney, invented a portable, non-invasive cap-like device that could translate thoughts into words when placed on the head. GrapheneX-UTS Human-centric Artificial Intelligence Centre Additionally, scientists from the University of Texas, Austin, developed to translate the thoughts in one's brain into actual speech by monitoring the brain's activity involving the use of fMRI and Al language models. a non-invasive technique The accuracy of these techniques was average. However, with the intended goal achieved, it would only take more years of development to increase its accuracy. Still, no one wants to walk around carrying a headpiece with cords sticking out, and neither can they carry fMRI machines all day. Hence, there is a need for less obvious, non-intrusive devices or less risky invasive BCIs. While the world can be swamped by Brainwave-controlled devices that read our minds, it poses serious concerns. One day, like every other day, we'll be carrying out activities with silent thoughts all in our heads, but this is the sad part: organizations would be able to read, monitor, store, access, and use these thoughts when needed. As tragic as this sounds, it is already in progress. The thoughts in our brain are currently our only inaccessible assets, our privacy safe, our most intimate thoughts, and our personal identity—no one should be able to access our minds. Data Privacy: But with the recent discoveries and technologies in neuroscience, this intrinsic privacy may be breached sooner than we know, leaving us truly vulnerable to Big Tech companies and organizations. If we lose control of our minds, if any company can spy into our thoughts, what else do we have left of ourselves? Just like we have on the internet, companies would be able to buy and access our brain data for targeted marketing. The first appeal for every product starts with the brain. You may not be able to say it, but you can think it accurately. Imagine advertisers having that exact information just as we feel it; then, they will be able to market their products in a practically irresistible way. Neurocomputing and mind-reading devices would make brain data available on a massive scale, and this is also very detrimental to us. Think about it; aside from compelling advertising, how safe are you? The brain is where all your secrets lie: your passwords, location, investments, amount of money you own, where you saved your assets, etc. A brain data hack would mean a loss too great to bear. The argument of some neuroscientists that the potential for misuse of neurocomputing and mind reading is so great that there needs to be a revamp of the human rights laws to protect us all is valid. Because this technology has what it takes to interfere with our most basic rights, the current laws are not fully equipped to salvage the situation; we need a new "jurisprudence of the mind" for our protection. So, in all, our biggest challenge is privacy and safety! What's the future? A quiet world. Maybe what we need? Or maybe not. Elon Musk and his company, Neuralink, have been at the forefront, pushing the concept of neurocomputing beyond the idea of it being the help for paraplegics only. For Musk, it would take conceptual interactions to a whole new level where words and conversations don't have to be verbalized, just thought. He calls it "consensual telepathy,” a realm where your speech becomes completely unnecessary. You can walk into a room with people conversing but not conversing because their thoughts are flying invisibly between them. Reading that now, it does sound uncanny. What becomes of our speech? Oh! I know what you are thinking, "there would never be a day when our speech isn't needed,” and of course, that's somewhat true. But come to think of it, before the advent of mobile phones, would there have been a day we thought we'd be so glued to these gadgets we'd rarely appreciate physical communication and rather be embedded in virtual conversations? I don't think so. When all you need is your thoughts to perform the most basic activities like ordering a coffee from a cafe, asking your friend to pass you the popcorn at a movie, communicating a proposal to your boss, conversing verbally on a date, etc., you may slowly lose touch of actual words. Almost the same way, we are slowly losing touch with writing on paper because we'd rather type into our computers, tablets, or phones. Well, there's no denying the fact that it sounds pretty interesting, too. You'd never have to say you don't know how to express yourself because your mind will do all the expressing for you where words would have escaped you. The concept of your feelings will flow just as you feel them, and no matter how disgruntled they are, your receiver will get the exact message with the fear, excitement, contempt, and all the fleeting emotions - conveyed just as it is. Also, the advancement of this technology would eliminate the language barrier. Whatever we think can be communicated directly to the recipient in whatever language desired. What neurocomputing proposes for the future is both unsettling and intriguing, but we really can't stop technology from evolving, and neither can we stop the future. We only hope that measures are instituted to mitigate the cons while we hope for the best.

From Science Fiction to Reality: The Promise of Neurocomputing and Brain-Computer Interfaces (BCIs)

কল্পবিজ্ঞান থেকে বাস্তবে: নিউরোকম্পিউটিং এবং ব্রেন-কম্পিউটার ইন্টারফেসের প্রতিশ্রুতি (BCIs)

Von der Science-Fiction zur Realität: Das Versprechen von Neurocomputing und Brain-Computer-Interfaces (BCIs)

De la ciencia ficción a la realidad: la promesa de la neurocomputación y las interfaces cerebro-computadora (BCI)

De la science-fiction à la réalité : la promesse de la neuroinformatique et des interfaces cerveau-ordinateur (BCI)

विज्ञान कथा से वास्तविकता तक: न्यूरोकंप्यूटिंग और मस्तिष्क-कंप्यूटर इंटरफेस (बीसीआई) का वादा

サイエンスフィクションから現実へ: ニューロコンピューティングと脳コンピュータインターフェース (BCI) の可能性

SF에서 현실로: 신경컴퓨팅과 뇌-컴퓨터 인터페이스(BCI)의 전망

Da ficção científica à realidade: a promessa da neurocomputação e das interfaces cérebro-computador (BCIs)

От научной фантастики к реальности: перспективы нейрокомпьютеров и интерфейсов «мозг-компьютер» (BCI)

Bilim Kurgudan Gerçeğe: Nörobilgisayar ve Beyin-Bilgisayar Arayüzlerinin (BCI) Vaadi

Từ khoa học viễn tưởng đến hiện thực: Lời hứa về giao diện máy tính thần kinh và não-máy tính (BCI)

从科幻到现实：神经计算和脑机接口 (BCI) 的前景

Last week I was talking about job applications and not holding yourself back by saying “no” pre-emptively. I started to dive into job descriptions before realizing they deserved their own post. Which is what I’m talking about today. I’m not going to sugar-coat it: Job descriptions are often a complete work of fiction. Sometimes, this is because the job description is years out of date – written before the previous employee in that role (or the one before  that) had even started. The original responsibilities morphed and  changed with the needs of the moment and the skills of the people, but the job description remained unchanged. Sometimes the job description started out realistically but then HR (or the “AI”-driven tool HR is using) decides to add other requirements  to better match similar job descriptions out of standardized databases –  which are themselves nothing more than amalgams of job descriptions  sourced across time and distance. It’s job-description-by-committee,  with no bearing on reality. Sometimes HR offers a single open req (when 3 are needed) so the manager tries to create a wish-list of requirements and  responsibilities, in the hope that someone will come along with most (or  at least many) of the boxes checked. It’s a pipe dream, but a  frustratingly common one. Regardless of the reason, the truth remains that many of the job descriptions out there are nothing more than a messy collection of vague  hopes, wishful thinking, and out-of-date ideas. Over three decades an nearly 20 jobs, I’ve lost count of the number of times I’ve shown up for an interview only to discover that the work  bore NO resemblance to the requirements on paper. What I find more  astonishing is that sometimes the manager knew this was the case, but  often they didn’t. The lesson in all of this goes back to the core message of the : In your job hunt, job is to apply. Don’t second guess whether you have “enough” of the  skills, because – as I’ve just explained – the work their need done may  be very different. last post your Do you like the company because of it’s culture, creation, community, calling, or some other aspect? Are you intrigued enough to want to find  out more? That’s all it takes. Apply. Also published . here

Job Descriptions and Their Fictional Hiring Expectations

919 Stories To Learn About Science Fiction

581 Stories To Learn About Non Fiction

671 Stories To Learn About Fiction

MARGARET’S FLITTIN’.

Disclaimer: The author works at Microsoft. The article is a general exploration of user trust and AI, but it does call out some references to the author’s company’s products such as Copilot/Azure Open AI. “I knew I had just seen the most important advance in technology since the graphical user interface.” — Bill Gates, about ChatGPT acing an AP-level biology test Generative AI’s surge in popularity has opened up new and innovative possibilities, transforming the way we live and work. According to a study conducted by PwC, AI could contribute up to $15.7 trillion to the global economy in 2030. A Forbes Advisor report shows that a staggering 97 percent of business owners believe ChatGPT will help their business. These are exciting times for new startups specializing in Generative AI. However, there is a dichotomy in how end users feel about the Generative AI revolution. For instance, 46 percent of Americans are equally concerned and excited about AI (Artificial Intelligence). This type of user mistrust and skepticism has the potential to encumber AI adoption. What are some ways to mitigate inaccurate or misleading content generated by Large Language Models? How can companies build products that users not only love but trust? Hallucinations Large Language Models (LLMs) have been labeled stochastic parrots — a fitting metaphor for entities that regurgitate convincingly plausible answers without really understanding the meaning behind them. As Stephen Wolfram put it, a model like ChatGPT is simply adding one word at a time. LLMs are trained on massive amounts of data from various sources but aren’t by themselves doing any lookups or referring to external sources of information. This means that an LLM can make misleading claims or factual errors — while sounding fairly confident about the mistake. These counterfactual statements are, somewhat anthropomorphically, nicknamed hallucinations. To some extent, LLMs are meant to hallucinate. It’s Generative AI’s version of “it’s not a bug — it’s by design.” The same capability that makes LLMs hallucinate is also what makes them powerful. On the flip side, there are situations when hallucinations are not desirable, and where it is critical to get truthful and factual responses. Working around the non-determinism of GenAI products continues to be a persistent challenge. But there are ways to ensure that we can still maintain user trust while harnessing all the creativity and potential behind the AI. Adding a Grounding Source of Truth One way to mitigate hallucinations is to augment the knowledge an LLM possesses with a grounding source of truth. Imagine you’re making someone take a test on a complex, previously unseen topic. This person is a fluent English speaker, but they know next to nothing about the topic they’re being tested on. Now, they could simply fabricate answers for all the questions. Or, you can give them an open textbook about the topic they’re being tested on, so they have reference material from which they can answer. This example is a simplified explanation of what grounding represents in the context of LLMs. Grounding is table stakes while creating any kind of custom LLM application, especially for internal data (such as customer service chatbots that rely on non–public-facing knowledge bases) or applications that rely on a corpus to retrieve answers (a.k.a., search engines). As far as implementation goes, Retrieval Augmented Generation is probably the most ubiquitous pattern currently used for grounding LLMs. There are numerous out-of-the-box RAG solutions in the market that take in a custom corpus to do retrieval and generation. Azure Open AI on your own data is one such example, where users can upload a dataset or document from blob storage and “chat” with it. The feature also offers citations to the source text, along with the response. Making an LLM cite an authoritative source as the reason behind a particular answer automatically increases transparency and adds credibility. Grounding can also help with evaluation and validation of response quality, a key component to testing Generative AI applications before making them production-worthy for public release. In the absence of a grounding dataset, there is no way to “compare” the responses of an LLM to see if they are based on truth or not. So, choosing the right implementation of RAG and the right grounding dataset is critical for improving the factualness of the GenAI application. Increasing Explainability Surveys have shown that people think providing too little context on AI decisions can erode trust. Deep Learning models are trained to get accurate and fast predictions, but the end users are often clueless as to why the model generated a particular response. A famous example of this black box behavior occurs in my favorite work of fiction, Douglas Adams’ brilliant Hitchhiker’s Guide to Galaxy. In the book, the supercomputer Deep Thought is tasked with working out the Answer to The Ultimate Question of Life, the Universe, and Everything. After 7.5 million years, Deep Thought responds that the answer is 42. The scientists are puzzled, but Deep Thought is quite confident about the answer: “I think the problem, to be quite honest with you, is that you’ve never actually known what the question was.” In order to then calculate the Ultimate Question, a special computer is built, and hilarities ensue — but I won’t spoil the book any further. Decades after the book was written, we now have ways to peek behind the curtain, to understand why an AI generated a particular response. While traditional Machine Learning (ML) models use techniques like LIME and SHAP for explainability, Chain-of-Thought (CoT) prompting is a popular technique to add more transparency to Generative AI’s reasoning process. In essence, CoT consists of meta-prompts asking the AI to explain itself or interpret why it arrived at a particular output. In fact, researchers have noted that just adding a phrase like “Let’s think step by step” also increases the accuracy of generated responses, even in zero-shot scenarios. Employing Guardrails Apart from grounding, there are also various other techniques to reduce hallucinations including advanced prompt engineering with CoVE (chain of thought verification), feedback loops with iterative improvements (or “humans in the loop”), using a combination of fine-tuned models with RAG, asking an LLM to self-critique its own output, as well as playing around with the “creativity” and “temperature” of LLM APIs. However, none of these methods are infallible since an LLM is essentially non-deterministic and unpredictable. There could be grave ethical as well as practical implications from Artificial Intelligence making decisions without any oversight. Incorrect predictions could lead to potentially life-threatening implications in fields like medical sciences, pharmacology, autonomous vehicle design, law enforcement, or disaster management. In the end, the safest option is for humans to have the final say in accepting or rejecting any AI-driven predictions. Having the end user in charge of vetting AI decisions also serves as a great feedback loop for developers to learn from and iterate continuously. Efforts to reduce hallucinations by improving factualness and transparency will increase user confidence in AI products. In addition, accountability is a key Responsible AI principle that must be at the center of all AI design. AI can be an amazing tool to revolutionize the way we all live and work — but ultimately, not to replace human judgment. Disclaimer: The author works at Microsoft. The article is a general exploration of user trust and AI, but it does call out some references to the author’s company’s products such as Copilot/Azure Open AI. Disclaimer: The author works at Microsoft. The article is a general exploration of user trust and AI, but it does call out some references to the author’s company’s products such as Copilot/Azure Open AI. “I knew I had just seen the most important advance in technology since the graphical user interface.” — Bill Gates, about ChatGPT acing an AP-level biology test “I knew I had just seen the most important advance in technology since the graphical user interface.” — Bill Gates, about ChatGPT acing an AP-level biology test “I knew I had just seen the most important advance in technology since the graphical user interface.” Bill Gates, about ChatGPT acing an AP-level biology test Bill Gates, about ChatGPT acing an AP-level biology test Generative AI’s surge in popularity has opened up new and innovative possibilities, transforming the way we live and work. According to a study conducted by PwC, AI could contribute up to $15.7 trillion to the global economy in 2030 . A Forbes Advisor report shows that a staggering 97 percent of business owners believe ChatGPT will help their business . These are exciting times for new startups specializing in Generative AI . AI could contribute up to $15.7 trillion to the global economy in 2030 97 percent of business owners believe ChatGPT will help their business new startups specializing in Generative AI However, there is a dichotomy in how end users feel about the Generative AI revolution. For instance, 46 percent of Americans are equally concerned and excited about AI (Artificial Intelligence). This type of user mistrust and skepticism has the potential to encumber AI adoption. What are some ways to mitigate inaccurate or misleading content generated by Large Language Models? How can companies build products that users not only love but trust? 46 percent of Americans are equally concerned and excited about Hallucinations Hallucinations Large Language Models (LLMs) have been labeled stochastic parrots — a fitting metaphor for entities that regurgitate convincingly plausible answers without really understanding the meaning behind them. As Stephen Wolfram put it, a model like ChatGPT is simply adding one word at a time. stochastic parrots adding one word at a time. LLMs are trained on massive amounts of data from various sources but aren’t by themselves doing any lookups or referring to external sources of information. This means that an LLM can make misleading claims or factual errors — while sounding fairly confident about the mistake. These counterfactual statements are, somewhat anthropomorphically , nicknamed hallucinations . somewhat anthropomorphically hallucinations To some extent, LLMs are meant to hallucinate. It’s Generative AI’s version of “it’s not a bug — it’s by design.” The same capability that makes LLMs hallucinate is also what makes them powerful. On the flip side, there are situations when hallucinations are not desirable, and where it is critical to get truthful and factual responses. meant Working around the non-determinism of GenAI products continues to be a persistent challenge. But there are ways to ensure that we can still maintain user trust while harnessing all the creativity and potential behind the AI. Adding a Grounding Source of Truth Adding a Grounding Source of Truth One way to mitigate hallucinations is to augment the knowledge an LLM possesses with a grounding source of truth. Imagine you’re making someone take a test on a complex, previously unseen topic. This person is a fluent English speaker, but they know next to nothing about the topic they’re being tested on. Now, they could simply fabricate answers for all the questions. Or, you can give them an open textbook about the topic they’re being tested on, so they have reference material from which they can answer. This example is a simplified explanation of what grounding represents in the context of LLMs. Grounding is table stakes while creating any kind of custom LLM application, especially for internal data (such as customer service chatbots that rely on non–public-facing knowledge bases) or applications that rely on a corpus to retrieve answers (a.k.a., search engines). As far as implementation goes, Retrieval Augmented Generation is probably the most ubiquitous pattern currently used for grounding LLMs. There are numerous out-of-the-box RAG solutions in the market that take in a custom corpus to do retrieval and generation. Retrieval Augmented Generation Azure Open AI on your own data is one such example, where users can upload a dataset or document from blob storage and “chat” with it. The feature also offers citations to the source text, along with the response. Making an LLM cite an authoritative source as the reason behind a particular answer automatically increases transparency and adds credibility. Azure Open AI on your own data Grounding can also help with evaluation and validation of response quality, a key component to testing Generative AI applications before making them production-worthy for public release. In the absence of a grounding dataset, there is no way to “compare” the responses of an LLM to see if they are based on truth or not. So, choosing the right implementation of RAG and the right grounding dataset is critical for improving the factualness of the GenAI application. Increasing Explainability Increasing Explainability Surveys have shown that people think providing too little context on AI decisions can erode trust . Deep Learning models are trained to get accurate and fast predictions, but the end users are often clueless as to why the model generated a particular response. providing too little context on AI decisions can erode trust why A famous example of this black box behavior occurs in my favorite work of fiction, Douglas Adams’ brilliant Hitchhiker’s Guide to Galaxy . Hitchhiker’s Guide to Galaxy In the book, the supercomputer Deep Thought is tasked with working out the Answer to The Ultimate Question of Life, the Universe, and Everything. After 7.5 million years, Deep Thought responds that the answer is 42. The scientists are puzzled, but Deep Thought is quite confident about the answer: “I think the problem, to be quite honest with you, is that you’ve never actually known what the question was.” “ I think the problem, to be quite honest with you, is that you’ve never actually known what the question was. ” I think the problem, to be quite honest with you, is that you’ve never actually known what the question was. question In order to then calculate the Ultimate Question, a special computer is built, and hilarities ensue — but I won’t spoil the book any further. Decades after the book was written, we now have ways to peek behind the curtain, to understand why an AI generated a particular response. While traditional Machine Learning (ML) models use techniques like LIME and SHAP for explainability, Chain-of-Thought (CoT) prompting is a popular technique to add more transparency to Generative AI’s reasoning process. In essence, CoT consists of meta-prompts asking the AI to explain itself or interpret why it arrived at a particular output. In fact, researchers have noted that just adding a phrase like “ Let’s think step by step ” also increases the accuracy of generated responses, even in zero-shot scenarios. Let’s think step by step Employing Guardrails Employing Guardrails Apart from grounding, there are also various other techniques to reduce hallucinations including advanced prompt engineering with CoVE (chain of thought verification), feedback loops with iterative improvements (or “humans in the loop”), using a combination of fine-tuned models with RAG, asking an LLM to self-critique its own output, as well as playing around with the “creativity” and “temperature” of LLM APIs. However, none of these methods are infallible since an LLM is essentially non-deterministic and unpredictable. There could be grave ethical as well as practical implications from Artificial Intelligence making decisions without any oversight. Incorrect predictions could lead to potentially life-threatening implications in fields like medical sciences, pharmacology, autonomous vehicle design, law enforcement, or disaster management. In the end, the safest option is for humans to have the final say in accepting or rejecting any AI-driven predictions. Having the end user in charge of vetting AI decisions also serves as a great feedback loop for developers to learn from and iterate continuously. Efforts to reduce hallucinations by improving factualness and transparency will increase user confidence in AI products. In addition, accountability is a key Responsible AI principle that must be at the center of all AI design. AI can be an amazing tool to revolutionize the way we all live and work — but ultimately, not to replace human judgment.

Pursuit of Trustworthiness: Increasing User Trust in Generative AI Products

This paper is available on arxiv under CC 4.0 license.
Authors:
(1) Anh Nguyen-Duc, University of South Eastern Norway, BøI Telemark, Norway3800 and Norwegian University of Science and Technology, Trondheim, Norway7012;
(2) Beatriz Cabrero-Daniel, University of Gothenburg, Gothenburg, Sweden;
(3) Adam Przybylek, Gdansk University of Technology, Gdansk, Poland;
(4) Chetan Arora, Monash University, Melbourne, Australia;
(5) Dron Khanna, Free University of Bozen-Bolzano, Bolzano, Italy;
(6) Tomas Herda, Austrian Post - Konzern IT, Vienna, Austria;
(7) Usman Rafiq, Free University of Bozen-Bolzano, Bolzano, Italy;
(8) Jorge Melegati, Free University of Bozen-Bolzano, Bolzano, Italy;
(9) Eduardo Guerra, Free University of Bozen-Bolzano, Bolzano, Italy;
(10) Kai-Kristian Kemell, University of Helsinki, Helsinki, Finland;
(11) Mika Saari, Tampere University, Tampere, Finland;
(12)  Zheying Zhang, Tampere University, Tampere, Finland;
(13) Huy Le, Vietnam National University Ho Chi Minh City, Hochiminh City, Vietnam and Ho Chi Minh City University of Technology, Hochiminh City, Vietnam;
(14) Tho Quan, Vietnam National University Ho Chi Minh City, Hochiminh City, Vietnam and Ho Chi Minh City University of Technology, Hochiminh City, Vietnam;
(15) Pekka Abrahamsson, Tampere University, Tampere, Finland. Table of Links Abstract and Introduction
Background
Research Approach
Research Agenda
Outlook and Conclusions
References Abstract Context Generative Artificial Intelligence (GenAI) tools have become increasingly prevalent in software development, offering assistance to various managerial and technical project activities. Notable examples of these tools include OpenAI’s ChatGPT, GitHub Copilot, and Amazon CodeWhisperer. Objective. Although many recent publications have explored and evaluated the application of GenAI, a comprehensive understanding of the current development, applications, limitations, and open challenges remains unclear to many. Particularly, we do not have an overall picture of the current state of GenAI technology in practical software engineering usage scenarios. Method. We conducted a literature review and focus groups for a duration of five months to develop a research agenda on GenAI for Software Engineering. Results. We identified 78 open Research Questions (RQs) in 11 areas of Software Engineering. Our results show that it is possible to explore the adoption of GenAI in partial automation and support decision-making in all software development activities. While the current literature is skewed toward software implementation, quality assurance and software maintenance, other areas, such as requirements engineering, software design, and software engineering education, would need further research attention. Common considerations when implementing GenAI include industry level assessment, dependability and accuracy, data accessibility, transparency, and sustainability aspects associated with the technology. Conclusions. GenAI is bringing significant changes to the field of software engineering. Nevertheless, the state of research on the topic still remains immature. We believe that this research agenda holds significance and practical value for informing both researchers and practitioners about current applications and guiding future research. Keywords: Generative Artificial Intelligence, GenAI, ChatGPT, CoPilot, Software Engineering, Software Project, Software Development, Focus Group, Literature review, Structured review, Literature survey, Research Agenda, Research Roadmap 1. Introduction In recent years, Generative Artificial Intelligence (GenAI) has attracted significant attention and interest from Software Engineering (SE) research and practice. GenAI tools such as GitHub Copilot1 and ChatGPT2 have been rapidly explored in various professional contexts given their remarkable performance in producing human-like content. The emerging adoption of  these tools reintroduces a host of classic concerns about productivity and quality when adopting new technologies. Clearly, code generation and test case optimization are SE tasks that directly benefit from recent large language models (LLMs) [1, 2, 3, 4]. Beyond traditional research on applied. Machine Learning (ML), GenAI tools introduced usability and accessibility, leveraging AI-generated content to a broader range of professionals, requiring less technical competence to work and integrate them into existing work environment. GenAI tools are also showing their potential in non-coding tasks, such as assisting requirements engineering, software processes and project management. At the moment, GenAI is an active research areas with several challenges and open questions. To do well on specific tasks, LLMs require fine-tuning or training. Research has much focused so far on achieving reliable and scalable GenAI output for different SE tasks. GenAI models are inherently nondeterministic: the same prompt produces different answers on different inference executions [5]. Moreover, GenAI’s output can be very sensitive to the input parameters or settings [6, 7]. Hallucination is another common concern with AI-generated content, as they can be imaginary or fictional [8]. The promise of AI automation might be closer than ever in SE [9], when these open challenges are sufficiently addressed. Research agenda is a popular type of work for guiding and organizing research efforts in certain research areas [10, 11, 12, 13, 14, 15]. It often includes a review of existing work, and a presentation of directions, visions and priorities, enabling researchers to make meaningful contributions to their respective fields and to address pressing challenges. This research agenda is driven by past and current work on GenAI for SE. Based on focus groups and a literature review, we portray the current research on GenAI in SE and present the research agenda with open challenges for future research. While we do not emphasize the comprehensiveness of the literature review due to the fast-moving nature of the topic, the focus groups present practical expectations of the future roles of GenAI in software development. The research agenda is organized into 11 areas of concern, which are: (1) Requirements Engineering, (2) Software Design,(3) Sofware Implementation, (4) Quality Assurance, (5) Software Maintenance and Evolution, (6) Software Processes and Tools, (7) Engineering Management, (8) Professional Competencies, (9) Software Engineering Education, (10) Macro Aspects and (11) Fundamental concerns of GenAI. Two international events have provided a ground for this work. As a part of the first international workshop on AI-assisted Agile Software Development3 , we ran our first focus group to explore the benefits and challenges of AI-assisted Agile software development. At the Requirements Engineering (RE) conference, we organized the second international workshop on Requirements Engineering for Software Startups and Emerging Technologies (RESET) 4 with a special theme on GenAI and Requirements Engineering. Consequently, the identified open challenges reflect not just the existing gaps in literature but also draw from the practical, value-driven insights of the co-authors. The paper is structured as follows: we clarify the meaning of GenAI and what we know about GenAI in SE in Section 2, and we will present our research approach in Section 3. Section 4 presents our Research Agenda. Finally, Section 5 is Outlook and Conclusions. This paper is available on arxiv under CC 4.0 license. Authors: (1) Anh Nguyen-Duc, University of South Eastern Norway, BøI Telemark, Norway3800 and Norwegian University of Science and Technology, Trondheim, Norway7012; (2) Beatriz Cabrero-Daniel, University of Gothenburg, Gothenburg, Sweden; (3) Adam Przybylek, Gdansk University of Technology, Gdansk, Poland; (4) Chetan Arora, Monash University, Melbourne, Australia; (5) Dron Khanna, Free University of Bozen-Bolzano, Bolzano, Italy; (6) Tomas Herda, Austrian Post - Konzern IT, Vienna, Austria; (7) Usman Rafiq, Free University of Bozen-Bolzano, Bolzano, Italy; (8) Jorge Melegati, Free University of Bozen-Bolzano, Bolzano, Italy; (9) Eduardo Guerra, Free University of Bozen-Bolzano, Bolzano, Italy; (10) Kai-Kristian Kemell, University of Helsinki, Helsinki, Finland; (11) Mika Saari, Tampere University, Tampere, Finland; (12)  Zheying Zhang, Tampere University, Tampere, Finland; (13) Huy Le, Vietnam National University Ho Chi Minh City, Hochiminh City, Vietnam and Ho Chi Minh City University of Technology, Hochiminh City, Vietnam; (14) Tho Quan, Vietnam National University Ho Chi Minh City, Hochiminh City, Vietnam and Ho Chi Minh City University of Technology, Hochiminh City, Vietnam; (15) Pekka Abrahamsson, Tampere University, Tampere, Finland. This paper is available on arxiv under CC 4.0 license. Authors: Authors: (1) Anh Nguyen-Duc, University of South Eastern Norway, BøI Telemark, Norway3800 and Norwegian University of Science and Technology, Trondheim, Norway7012; (2) Beatriz Cabrero-Daniel, University of Gothenburg, Gothenburg, Sweden; (3) Adam Przybylek, Gdansk University of Technology, Gdansk, Poland; (4) Chetan Arora, Monash University, Melbourne, Australia; (5) Dron Khanna, Free University of Bozen-Bolzano, Bolzano, Italy; (6) Tomas Herda, Austrian Post - Konzern IT, Vienna, Austria; (7) Usman Rafiq, Free University of Bozen-Bolzano, Bolzano, Italy; (8) Jorge Melegati, Free University of Bozen-Bolzano, Bolzano, Italy; (9) Eduardo Guerra, Free University of Bozen-Bolzano, Bolzano, Italy; (10) Kai-Kristian Kemell, University of Helsinki, Helsinki, Finland; (11) Mika Saari, Tampere University, Tampere, Finland; (12)  Zheying Zhang, Tampere University, Tampere, Finland; (13) Huy Le, Vietnam National University Ho Chi Minh City, Hochiminh City, Vietnam and Ho Chi Minh City University of Technology, Hochiminh City, Vietnam; (14) Tho Quan, Vietnam National University Ho Chi Minh City, Hochiminh City, Vietnam and Ho Chi Minh City University of Technology, Hochiminh City, Vietnam; (15) Pekka Abrahamsson, Tampere University, Tampere, Finland. Table of Links Abstract and Introduction Background Research Approach Research Agenda Outlook and Conclusions References Abstract and Introduction Abstract and Introduction Background Background Research Approach Research Approach Research Agenda Research Agenda Outlook and Conclusions Outlook and Conclusions References References Abstract Context Generative Artificial Intelligence (GenAI) tools have become increasingly prevalent in software development, offering assistance to various managerial and technical project activities. Notable examples of these tools include OpenAI’s ChatGPT, GitHub Copilot, and Amazon CodeWhisperer. Context Objective. Although many recent publications have explored and evaluated the application of GenAI, a comprehensive understanding of the current development, applications, limitations, and open challenges remains unclear to many. Particularly, we do not have an overall picture of the current state of GenAI technology in practical software engineering usage scenarios. Objective. Method . We conducted a literature review and focus groups for a duration of five months to develop a research agenda on GenAI for Software Engineering. Method Results . We identified 78 open Research Questions (RQs) in 11 areas of Software Engineering. Our results show that it is possible to explore the adoption of GenAI in partial automation and support decision-making in all software development activities. While the current literature is skewed toward software implementation, quality assurance and software maintenance, other areas, such as requirements engineering, software design, and software engineering education, would need further research attention. Common considerations when implementing GenAI include industry level assessment, dependability and accuracy, data accessibility, transparency, and sustainability aspects associated with the technology. Results Conclusions. GenAI is bringing significant changes to the field of software engineering. Nevertheless, the state of research on the topic still remains immature. We believe that this research agenda holds significance and practical value for informing both researchers and practitioners about current applications and guiding future research. Conclusions. Keywords: Generative Artificial Intelligence, GenAI, ChatGPT, CoPilot, Software Engineering, Software Project, Software Development, Focus Group, Literature review, Structured review, Literature survey, Research Agenda, Research Roadmap Keywords: 1. Introduction In recent years, Generative Artificial Intelligence (GenAI) has attracted significant attention and interest from Software Engineering (SE) research and practice. GenAI tools such as GitHub Copilot1 and ChatGPT2 have been rapidly explored in various professional contexts given their remarkable performance in producing human-like content. The emerging adoption of  these tools reintroduces a host of classic concerns about productivity and quality when adopting new technologies. Clearly, code generation and test case optimization are SE tasks that directly benefit from recent large language models (LLMs) [1, 2, 3, 4]. Beyond traditional research on applied. Machine Learning (ML), GenAI tools introduced usability and accessibility, leveraging AI-generated content to a broader range of professionals, requiring less technical competence to work and integrate them into existing work environment. GenAI tools are also showing their potential in non-coding tasks, such as assisting requirements engineering, software processes and project management. At the moment, GenAI is an active research areas with several challenges and open questions. To do well on specific tasks, LLMs require fine-tuning or training. Research has much focused so far on achieving reliable and scalable GenAI output for different SE tasks. GenAI models are inherently nondeterministic: the same prompt produces different answers on different inference executions [5]. Moreover, GenAI’s output can be very sensitive to the input parameters or settings [6, 7]. Hallucination is another common concern with AI-generated content, as they can be imaginary or fictional [8]. The promise of AI automation might be closer than ever in SE [9], when these open challenges are sufficiently addressed. Research agenda is a popular type of work for guiding and organizing research efforts in certain research areas [10, 11, 12, 13, 14, 15]. It often includes a review of existing work, and a presentation of directions, visions and priorities, enabling researchers to make meaningful contributions to their respective fields and to address pressing challenges. This research agenda is driven by past and current work on GenAI for SE. Based on focus groups and a literature review, we portray the current research on GenAI in SE and present the research agenda with open challenges for future research. While we do not emphasize the comprehensiveness of the literature review due to the fast-moving nature of the topic, the focus groups present practical expectations of the future roles of GenAI in software development. The research agenda is organized into 11 areas of concern, which are: (1) Requirements Engineering, (2) Software Design,(3) Sofware Implementation, (4) Quality Assurance, (5) Software Maintenance and Evolution, (6) Software Processes and Tools, (7) Engineering Management, (8) Professional Competencies, (9) Software Engineering Education, (10) Macro Aspects and (11) Fundamental concerns of GenAI. Two international events have provided a ground for this work. As a part of the first international workshop on AI-assisted Agile Software Development3 , we ran our first focus group to explore the benefits and challenges of AI-assisted Agile software development. At the Requirements Engineering (RE) conference, we organized the second international workshop on Requirements Engineering for Software Startups and Emerging Technologies (RESET) 4 with a special theme on GenAI and Requirements Engineering. Consequently, the identified open challenges reflect not just the existing gaps in literature but also draw from the practical, value-driven insights of the co-authors. The paper is structured as follows: we clarify the meaning of GenAI and what we know about GenAI in SE in Section 2, and we will present our research approach in Section 3. Section 4 presents our Research Agenda. Finally, Section 5 is Outlook and Conclusions.

FLIPTION

fliption.com

-- employees

Since n.d.

COMPANY RANKING

FLIPTION

EVERGREEN INDEX #4367

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

Fliption WIKI

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

FLIPTION

fliption.com

-- employees

Since n.d.

COMPANY RANKING

FLIPTION

EVERGREEN INDEX #4367

RELATED COMPANIESRANK

STORIES ABOUT FLIPTION

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

Fliption WIKI

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