Bugge Holm Hansen

It’s time to talk about Ai ethics Amid rumors that Open AI has used the voice of Scarlet Johansson, the debate regarding deep fake and the ethical side of technology use has gained more traction. OpenAI’s CEO Sam Altman suggested that the vocal design was intentionally mimicking Johansson’s character, posting a one-word tweet after the presentation that simply said “her”. Businesses must proactively address responsible AI practices before regulations come into force, such as the impending EU AI Act, that will come into effect next month. AI ethics and compliance are poised to become integral components of a business just like ESG. #AIethics #OpenAI #ChatGPT #ethics #tech #euaiact #technology #openai

29

1 Kommentar

🌍 Introducing the Verifiable Educational ID for European Students and Staff! We are thrilled to announce the Verifiable Educational ID—a revolutionary step for educational use cases in Europe. This EBSI schema ensures a secure and standardized digital identity for students and staff, enabling seamless interactions and access to services across institutions. What is the Verifiable Educational ID? 📄 Technical Details: - Type: Object - Credential Subject: Includes unique identifiers (DID:Key, eduPersonPrincipalName, etc.) - Schema Format: urn:schac:personalUniqueCode:int:esi:<sHO>:<code> Benefits: 1. Enhanced Mobility: Facilitates easy student and staff exchanges across educational institutions, bolstering the European Education Area. 2. Unified Access: Grants streamlined access to shared services, such as libraries and digital resources. 3. High Security: Ensures robust data protection and interoperability, adhering to EU standards. Why it Matters: This initiative empowers educational institutions to collaborate more effectively, aligning with the EU's digital transformation goals and fostering innovation and inclusivity in higher education. Learn more about how this schema can transform the educational experience in Europe! #HigherEducation #DigitalTransformation #EBSI #EuropeanUnion #EdTech #StudentMobility #EducationalInnovation #DC4EU

42

1 Kommentar

Finding the right use cases is key to utilizing #AI effectively. 

2

I have always argued that the obsession with Artificial General Intelligence (AGI) is misguided. If you have several narrow AI models that are very good at coordinating work, it will do pretty much the same for a large number of applications. This is exactly what OpenAI's 'Swarm Framework' is doing. Read more about this perspective: https://lnkd.in/eDSceJWG #aiinnovation #openai #aialignment #agi #artificialgeneralintelligence

1

How to prevent AI-powered world destruction: The clash between Innovation and Regulation As a Silicon Valley nomad in Norway, I have a lot of discussions on ‘need for AI regulation’ and importance of the EU AI act vs the ‘less restrictive’  movements in CA. While I agree that ‘today’ AI development, integration AND use in the valley and across US is a bit of ‘wild west’ open field, I see positives and negatives to this. First, we must recognize that ‘commercial B2C’ AI tools are still in their infancy. Innovation – in development and usage is critical – within safe boundaries. Therefore, I believe the role of regulations is to provide guidelines. People need then need to use their human intuition, critical thinking, knowledge of society to then make the right decisions, take the most responsible actions when using the tools – within the guidelines. Lets keep in mind that during early adoption stages  the ‘world wide web’, mobile apps etc each grew through a similar period of ‘lets try it’. During  this period of rapid innovation, business models, use cases and then regulations were developed for the technology to then fuel societal growth. I believe the same will occur with AI. In my typical Silicon Valley style, I believe BOTH innovation and regulation are important but they must include HUMAN ingenuity to ensure AI is best utilized to enhance society, the environment and people’s worklife.  Elin HaugeSarfraz M Paz (AM.IOD)Indira B.Thomas F. Anglero Barry Jones Cosimo Turroturro https://lnkd.in/dc2thcJ2

6

A study by the BCG Henderson Institute illuminates a pivotal shift: Generative AI (GenAI) is not just a productivity booster—it's redefining the capabilities of the modern workforce. The study aims to guide business leaders in understanding the symbiotic relationship between humans and GenAI in the future workplace. It reveals that a new breed of knowledge worker, enhanced by GenAI, is emerging on the global stage. These workers are not only writing code at unprecedented speeds but also creating personalized marketing content from single prompts and summarizing vast arrays of documents in mere seconds. This leads us to a critical insight: Leaders should view GenAI not merely as a tool for enhancing productivity but as a transformative technology that prepares the workforce for the evolving demands of today’s and tomorrow’s jobs. They should consider generative AI an exoskeleton: a tool that empowers workers to perform better, and do more, than either the human or GenAI can on their own. In other words. It's time for leaders to rethink how technology empowers future teams and redefine what is possible in the workplace. #GenerativeAI #FutureOfWork #Leadership #Innovation

102

5 Kommentarer

🚀🔍 𝐃𝐢𝐝 𝐲𝐨𝐮 𝐤𝐧𝐨𝐰 𝐭𝐡𝐚𝐭 𝐧𝐨𝐭 𝐚𝐥𝐥 '𝐨𝐩𝐞𝐧 𝐬𝐨𝐮𝐫𝐜𝐞' 𝐀𝐈 𝐦𝐨𝐝𝐞𝐥𝐬 𝐚𝐫𝐞 𝐚𝐬 𝐨𝐩𝐞𝐧 𝐚𝐬 𝐭𝐡𝐞𝐲 𝐜𝐥𝐚𝐢𝐦 𝐭𝐨 𝐛𝐞? 𝐴 𝑟𝑒𝑐𝑒𝑛𝑡 𝑠𝑡𝑢𝑑𝑦 ℎ𝑎𝑠 𝑠ℎ𝑒𝑑 𝑙𝑖𝑔ℎ𝑡 𝑜𝑛 𝑡ℎ𝑖𝑠 𝑖𝑚𝑝𝑜𝑟𝑡𝑎𝑛𝑡 𝑖𝑠𝑠𝑢𝑒, 𝑟𝑒𝑣𝑒𝑎𝑙𝑖𝑛𝑔 𝑡ℎ𝑎𝑡 𝑚𝑎𝑛𝑦 𝑏𝑖𝑔 𝑡𝑒𝑐ℎ 𝑐𝑜𝑚𝑝𝑎𝑛𝑖𝑒𝑠 𝑎𝑟𝑒 𝑢𝑠𝑖𝑛𝑔 𝑡ℎ𝑒 𝑡𝑒𝑟𝑚 '𝑜𝑝𝑒𝑛 𝑠𝑜𝑢𝑟𝑐𝑒' 𝑟𝑎𝑡ℎ𝑒𝑟 𝑙𝑜𝑜𝑠𝑒𝑙𝑦, 𝑟𝑒𝑠𝑡𝑟𝑖𝑐𝑡𝑖𝑛𝑔 𝑎𝑐𝑐𝑒𝑠𝑠 𝑡𝑜 𝑐𝑟𝑖𝑡𝑖𝑐𝑎𝑙 𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑠 𝑙𝑖𝑘𝑒 𝑐𝑜𝑑𝑒 𝑎𝑛𝑑 𝑡𝑟𝑎𝑖𝑛𝑖𝑛𝑔 𝑑𝑎𝑡𝑎. 😲 The study, conducted by Mark Dingemanse and Andreas Liesenfeld, highlights a practice known as "open-washing," where companies present their AI models as open source but limit transparency to maintain control and reap public relations benefits. Interestingly, it’s the smaller players, not the tech giants, who are truly leading the charge in AI transparency. 📊✨ 𝐊𝐞𝐲 𝐓𝐚𝐤𝐞𝐚𝐰𝐚𝐲𝐬: 1️⃣ 𝐌𝐢𝐬𝐥𝐞𝐚𝐝𝐢𝐧𝐠 𝐎𝐩𝐞𝐧𝐧𝐞𝐬𝐬: Many AI models from big tech companies are not fully open, restricting access to code, training data, and model weights. 2️⃣ 𝐎𝐩𝐞𝐧-𝐖𝐚𝐬𝐡𝐢𝐧𝐠: Companies like Meta and Microsoft claim openness but often fall short, revealing minimal details. 3️⃣ 𝐒𝐦𝐚𝐥𝐥 𝐏𝐥𝐚𝐲𝐞𝐫𝐬 𝐒𝐡𝐢𝐧𝐞: Smaller firms and academic collaborations, like those behind the BLOOM model, tend to be more transparent and genuinely open. 4️⃣ 𝐄𝐔 𝐀𝐈 𝐀𝐜𝐭 𝐈𝐦𝐩𝐚𝐜𝐭: The upcoming EU Artificial Intelligence Act will place significant legal weight on the definition of open source, influencing how models are regulated. 5️⃣ 𝐒𝐜𝐢𝐞𝐧𝐜𝐞 𝐚𝐧𝐝 𝐑𝐞𝐩𝐫𝐨𝐝𝐮𝐜𝐢𝐛𝐢𝐥𝐢𝐭𝐲: Full transparency is crucial for scientific progress and reproducibility, allowing researchers to understand and build upon existing models. This discussion is crucial as we move towards a more transparent and accountable AI future. What do you think about the current state of AI model transparency? Should big tech companies be more open about their AI technologies? 🤔💬 🔗 For a deeper dive into this topic, check out the full article on Nature’s website here: https://lnkd.in/dwuz8YyR 𝐘𝐨𝐮𝐫 𝐭𝐡𝐨𝐮𝐠𝐡𝐭𝐬 𝐚𝐧𝐝 𝐜𝐨𝐦𝐦𝐞𝐧𝐭𝐬 𝐚𝐫𝐞 𝐡𝐢𝐠𝐡𝐥𝐲 𝐯𝐚𝐥𝐮𝐞𝐝! Let’s discuss the future of open-source AI and its implications for science and technology. 🌐🔧 💡 𝐑𝐞𝐚𝐜𝐭, 💬 𝐂𝐨𝐦𝐦𝐞𝐧𝐭 and ♻️ 𝐑𝐞𝐩𝐨𝐬𝐭 to spread the knowledge. Follow Benny Langstrom and hit the 🔔 to be sure to catch my posts. #ai #technology #innovation

19

65 Kommentarer

OpenAI is set to launch GPT-4o mini, a more compact version of its advanced AI model, later today. The model (which the company calls “the most capable and cost-efficient small model available today”) is designed for high efficiency and low cost, with planned future enhancements to support image, video, and audio integrations. The GPT-4o mini is derived from the GPT-4o, which is notable for its ability to process content in 50 languages and support multiple media types at enhanced speeds and quality. This expansion is part of OpenAI’s strategy to lead in the multimodal AI space, where applications like ChatGPT can handle all media formats seamlessly within a single tool. GPT-4o mini will initially be available to free ChatGPT users, as well as ChatGPT Plus and Team subscribers, and it will roll out to Enterprise users next week. There are compelling reasons to shrink LLMs; smaller models should require smaller amounts of compute, which should translate to smaller amounts of electricity. We’ll see. -s

13

1 Kommentar

The rising cost of training models, saturation of benchmarking metrics, high incidence of hallucinations/fabrications, low scores on abstract reasoning, and high distrust by coders suggest AI isn’t working as well as the hype suggests.   The message from co-directors is positive: “A decade ago, the best AI systems in the world were unable to classify objects in images at a human level. AI struggled with language comprehension and could not solve math problems. Today, #AI systems routinely exceed human performance on standard benchmarks. Progress accelerated in 2023. New state-of-the-art systems like GPT-4, Gemini, and Claude 3 are impressively multimodal: They can generate fluent text in dozens of languages, process audio, and even explain memes. As AI has improved, it has increasingly forced its way into our lives. Companies are racing to build AI-based products, and AI is increasingly being used by the general public. But current AI #technology still has significant problems. It cannot reliably deal with facts, perform complex reasoning, or explain its conclusions.”   The last sentence points to problems such as:   “Figure 1.3.22 visualizes the training cost of all AI models for which the AI Index has estimates. As the figure shows, model training costs have sharply increased over time.” The figure shows that the training costs have risen about 1,000 times in last three years, and now exceed $100 million.   An emerging theme is “observed saturation on many benchmarks, such as ImageNet, used to assess the proficiency of AI models. Performance on these benchmarks has stagnated in recent years.” Past State of AI reports showed the exponentially rising costs for increasing image recognition accuracy from 90% to 99% and 99.9%. Those exponentially rising costs have likely led to the current saturation.   “#Research indicates that ChatGPT fabricates unverifiable information in approximately 19.5% of its responses, with these fabrications spanning a variety of topics such as language, climate, and technology. Furthermore, the study also reveals “that many LLMs struggle with identifying #hallucinations.”   “On many coding tasks, AI models are challenged to generate usable code or to solve computer science problems.” Only 2.85% of coding survey respondents highly trust AI, less than the 5.46% who highly distrust AI.   “The study revealed that GPT-4 significantly trails behind humans in abstract reasoning abilities: While humans score 95% on the benchmark, the best GPT-4 system only scores 69%.”   To achieve rapid market growth, we need reductions in frequency of hallucinations, re-acceleration of benchmarking metrics, and highly accurate code, none of which are occurring right now. The one area that impressed me was improvements in generating “a hyper-realistic image of Harry Potter.” #innovation #startups #artificialintelligence https://lnkd.in/gkAbR5xw

189

73 Kommentarer

Our report on #generativeai #systems is available. The work on Gen AIS is ongoing and will end in June with our last plenary meeting followed with the publication of the final report and a list of best practices. The aim of this work is to make it easier for public and private #organizations to sort out the positive and negative #impacts of generative #ai systems and to help them identify and focus on the relevant ones. A huge amount of work has been carried out over the last 5 months to identify, map, organise and prioritise the impacts of generative AIS, culminating in the production of three deliverables enabling organisations to optimise their decision-making in terms of the development, integration, deployment and use of these technologies.

22

3 Kommentarer

The Midas Plague Of Artificial Intelligence = This is a sophisticated post on Late Stage Capitalism interpreted as: Meaningless Production to support Meaningless Consumption to benefit less than 1% of Humanity We start with: Midas World is a collection of science fiction short stories by American writer Frederik Pohl, published in 1983. We focus on the following story: "The Midas Plague" (originally published in Galaxy in 1954). In a world of cheap energy, robots are overproducing the commodities enjoyed by humankind. The lower-class "poor" must spend their lives in frantic consumption, trying to keep up with the robots' extravagant production, while the upper-class "rich" can live lives of simplicity. Property crime is nonexistent, and the government Ration Board enforces the use of ration stamps to ensure that everyone consumes their quotas. The story deals with Morey Fry, who marries a woman from a higher-class family. Raised in a home with only five rooms she is unused to a life of forced consumption in their mansion of 26 rooms, nine automobiles, and five robots, causing arguments. Trained as an engineer, Morey modifies his robots to enjoy helping to consume his family's quota. He fears punishment when his idea is discovered, but the Ration Board—which has been looking for a way to abolish itself—quickly implements Morey's idea across the world. = Given this introduction let's now tackle the topic of this post: The Midas Plague Of Artificial Intelligence Here’s a guy: Damian Williams the United States Attorney for the Southern District of New York, and Christie M. Curtis, the Acting Assistant Director in Charge of the New York Field Office of the Federal Bureau of Investigation (FBI), announced the unsealing of a three-count criminal Indictment charging MICHAEL SMITH in connection with a scheme to create hundreds of thousands of songs with artificial intelligence and use automated programs called “bots” to stream the AI-generated songs billions of times. SMITH fraudulently obtained more than $10 million in royalty payments through his scheme. SMITH was arrested and will be presented before a U.S. Magistrate Judge in North Carolina. … SMITH created thousands of accounts on the Streaming Platforms (the “Bot Accounts”) that he could use to stream songs. He then used software to cause the Bot Accounts to continuously stream songs that he owned. At a certain point in the charged time period, SMITH estimated that he could use the Bot Accounts to generate approximately 661,440 streams per day, yielding annual royalties of $1,207,128. There was a famous article: Software is eating the world Never in my wildest imagination I thought of this! = Now we may be in a position to understand how Artificial Intelligence may facilitate both limitless Production and limitless Consumption to benefit less than 1% of humanity: That's the Midas Plague of Late Stage Capitalism Are you prepared for this Brave New World of Money For Nothing?

34

14 Kommentarer

Embracing generative AI tools for crafting PowerPoint presentations is a smart move. These tools are not just time-savers; they’re like a creativity booster shot for your slide decks! Ready to give your presentations a professional polish with a little AI magic? https://lnkd.in/gs2JPQQY

13

The Future of AI: Safe Superintelligence (SSI) Raises $1 Billion In a groundbreaking move, Safe Superintelligence (SSI), a new AI startup co-founded by former #OpenAI chief scientist Ilya Sutskever, has raised a whopping $1 billion in funding to develop safe #AI systems that surpass human intelligence. The Mission: #SSI is dedicated to prioritizing AI safety, focusing years on R&D to create superintelligent AI systems that benefit humanity. The Founders: Ilya Sutskever, Daniel Gross, and Daniel Levy founded SSI just three months ago, and the company is now valued at $5 billion. The Plan: With only ten employees, SSI will use the funds to acquire computing power and attract top talent. The Investors: Major venture capital firms like Andreessen Horowitz and Sequoia Capital are backing SSI. Why it Matters: This historic seed round marks a significant milestone in AI development. With Ilya Sutskever at the helm, SSI is poised to revolutionize the industry and attract exceptional talent. Let's discuss: What does this mean for the future of AI? Share your thoughts! https://lnkd.in/e_Y_2KQk?

1

1 Kommentar

It's a pleasure to serve on the editorial advisory board of the MIT Sloan Management Review. Among other things, my colleagues, expert panelists around the world, and I receive questions about #AI trends. The latest question? Whether we agree or disagree with the statement that "organizations will be ready to meet the requirements of the EU AI Act as they phase in over the next 12 months." The experts' response? Nearly half disagree or strongly disagree, while one-third neither agree nor disagree, and only one-fifth agree or strongly agree. My point of view: "Organizations will have no choice but to meet most of these requirements ... eventually. And not all requirements are created equal. Some requirements, like a high level of cybersecurity, will lead to immediate action. Other requirements, such as high-quality data sets or detailed documentation, have less dire consequences if not implemented immediately or consistently. One of the hardest requirements to meet would be that for transparency of AI algorithms. I am a general supporter of the EU AI Act, and understand the reasons for the transparency requirement in order to discern judgment, reasoning, and decision-making by machines. But the reality is, we do not have that transparency into human decision-making either. AI will get even more complex, and I fear this requirement is unattainable. There could be other ways to minimize bias and maximize fairness, including diversity of data sets; implementation of checks and balances; the proverbial humans in the loop; stress-testing that scrutinizes the levels of accuracy, fairness, and bias in outcomes; and validation by additional machine systems. Read more and share what you think: https://lnkd.in/ei_Pu-_e

55

5 Kommentarer

Why OpenAI's Safety Team Imploded It is very concerning to me that OpenAI's #AI Safety Team has disbanded because the Team's leadership felt OpenAI was more focused on shipping cool AI products vs. preventing serious long-term harm from AI. It's time for some reflection in the AI industry as well as external reviews of what Big Tech is doing in this area. https://lnkd.in/e45CwVwj

4

3 Kommentarer

A compelling example illustrating the potential challenges facing VR technology's widespread acceptance is the case of #3D TVs, which failed to gain traction in the market. One key factor contributing to their lack of popularity was the limited availability of #content, similar to the current dilemma facing VR. Additionally, discomfort caused by prolonged use of 3D glasses deterred many potential users, highlighting the importance of user comfort and accessibility in adopting new #technologies. Here is a great article on this topic written by Professor Scott Galloway: "#VR is a decade-long experiment that has cost tens of billions of dollars to prove nobody wants it. Consumers putting something on their face designed by Stanford and Harvard engineering graduates who live in Redmond or San Jose is less likely than the next great SaaS company emerging from Florence." https://lnkd.in/d9d3MJ2H

3

Although I haven't yet updated my paper, "Adding a 'Group Advisory Layer' to Your Use of Generative AI Tools Through Structured Prompting: Using Personas for Advisory Boards, Task Forces, Mastermind Groups, and Other Collections of Personas to Assist in Evaluations, Assessments, Recommendations, Decision-making, and much more (Including Law-related Examples)," I keep returning to and expanding the ideas in the paper I wrote a year ago. If I am known for anything in the realm of AI (other than my teaching of AI to law students), I'd like it to be for the core ideas from this paper and the applications arising out of them. Lately, I've included approaches with simulations (including simulations of simulations results), group personas, and approaches like "second pair of eyes," among other things. This area is such a rich vein for me to mine. I encourage you to experiment with it. Expect to hear more from me on this topic in the near future. #generativeai #personas #mastermindgroups #groupadvisorylayer #advisoryboards https://lnkd.in/gUEp6eVq Caitlin (Cat) Moon Nicole Morris April Dawson Sarah Freye Michele Jackson Bridgette Carr Bridget Mary McCormack Jenny Carter-Johnson Dennis Garcia ⚖️ Maya Markovich Thomas Officer Michael Cappucci Amy J Brookbanks Amy Conroy Caitlin McCabe Kristen Sonday Martin Schwimmer Patrick J. McKenna Allison C. Johs Elise AuxierCPC,PCC Michael Kraft Thomas G. Martin Nicola Shaver Nicole Black Marc Lauritsen Natalie Anne K. Li Jen Chang Michael Suttle Emmi Bane Richard Jolly Andy Unger Brian W Tang Mitch Jackson, Esq. Jacqueline Schafer Michael Grazio JoAnn Stonier Felix Marx Johnny Wong Alan Drimmer Ross Dawson

23

3 Kommentarer

"When it comes to the likes of the Humane AI Pin and Rabbit R1, one question is common: Why is this a separate device? The core functionality could just as easily be a smartphone app (and, in Rabbit’s case, it kind of is)." My take: I think it's too early to dismiss these devices. If anything, the Humane AI Pin can be the first step to replacing smart phones. Imagine a wearable pin that connects to smart glasses that are able to accept verbal commands and provide auditory and visual data through the glasses. What do you think?

1