Eye On A.I.

This episode is sponsored by Netsuite by Oracle, the number one cloud financial system, streamlining accounting, financial management, inventory, HR, and more. Download NetSuite's popular KPI Checklist, designed to give you consistently excellent performance - absolutely free at NetSuite.com/EYEONAI On Episode #137 of Eye on AI, we dive deep into the world of data storage and its connection to AI. Join host Craig Smith as we unravel valuable insights shared by Raul Martynek, CEO of Data Bank, and explore the evolving landscape of data centers and their pivotal role in powering AI-driven innovations. In this episode we discuss the history and significance of data centers, from their humble beginnings to their crucial role in today's data-driven world. Uncover how AI is driving the surging demand for data center capacity and the challenges the industry faces, including the impact of inflation and supply chain disruptions. We talk about China's ambitious plans to bolster its semiconductor industry and its implications for data center growth. We also delve into the critical issue of power consumption driven by AI and its effects on global power grids Join us for a thought-provoking discussion on the future of data centers and its synergy with AI. Raul Martynek's LinkedIn: https://www.linkedin.com/in/raulmartynek Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI (00:00) Preview and Introduction (01:43) Netsuite by Oracle (06:10) Growing Demands for Data Centers (12:35) Data Centres Before AI (24:08) Industry Challenges and Market Trends (29:20) AI's Impact on Power Consumption (36:00) Future of Databank (40:20) AIs Impact on Data Centers (44:34) Podcast Wrap-Up and Netsuite by Oracle

This episode is sponsored by MindStudio by YouAi. MindStudio is the best way to build an AI business. Start driving some serious revenue before everyone else. Mind Studio allows you to use conversational language to program incredibly powerful AI tools. No coding knowledge is needed to start your AI business. Sign up - YOUAI.AI/mindstudio On episode #135 of the Eye on AI podcast, Craig Smith sits down with Diaa El All, founder and CEO of Soundful, an AI tool that generates royalty free background music at the click of a button for your videos, streams, podcasts and more. From its integration with natural language processing to major industry partnerships, Soundful is paving the way for creators to not just produce, but also to monetize their content without worrying about being copyright striked. As we venture into the future of music production and licensing, we question the impact of AI on the music industry and what it means on the business side as well. We breakdown how AI can enable musicians to boost their creativity by drawing inspiration from generative AI music. Finally, we delve into the debate surrounding the role of imperfections in crafting a successful song. Imperfections have long been considered an integral part of what makes music human and relatable. Is perfection still the goal, or is there room for quirks and idiosyncrasies in AI-generated music? Tune in as we unravel the transformative power of AI in reshaping the music industry and propelling its evolution. Diaa El All LinkedIn: https://www.linkedin.com/in/diaaelall/ Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI (00:00) Preview and Introduction (01:36) MindStudio by YouAI (03:53) Soundful's Origin Story (06:25) Can Creators Benefit From AI Music? (16:24) Who uses Soundful? (25:50) Future of Music Production and Licensing (33:16) How AI Creates Music (37:15) MindStudio by YouAI

This episode is sponsored by Shopify. Shopify is a commerce platform that allows anyone to set up an online store and sell their products. It's the leading commerce platform designed for a business of any size. Whether you're selling online, on social media, or in person, Shopify has you covered on every base. With Shopify you can sell physical and digital products. You can sell services, memberships, ticketed events, rentals and even classes and lessons. Sign up for a $1 per month trial period at shopify.com/eyeonai On episode #134 of the Eye on AI podcast, Craig Smith sits down with Alex Zhavoronkov, founder and CEO of Insilico Medicine. Being at the forefront of cutting-edge drug discovery and longevity, Alex leverages the power of AI to develop novel drugs that could potentially extend our lifespan. In this episode we explore the unique tools Insilico Medicine uses to hypothesize protein targets and their links to diseases. Alex also gives us a peek into their automated robotic lab in Suzhou, China, which is expected to revolutionize drug development. We delve deep into the world of drug discovery, with a special focus on Insilico Medicine's tools, Pandomics, and Chemistry42, that are reshaping the field. Yet beyond the science, we also discuss the broader implications of tackling aging - from overpopulation to the strain on healthcare and social security systems. Finally, we tackle the elephant in the room - the challenges in the pharmaceutical industry. Can AI expedite the drug discovery process? Alex certainly thinks so. We explore how AI can help identify partners quickly, streamline processes, and ultimately accelerate the development of drugs. Join us as we unravel these mysteries and take you through the fascinating world of biotech, AI, and pharmaceutical research. (0:00) Preview (01:36) Shopify (04:26) AI's Revolutionizing Drug Discovery (06:11) Inside AI Pharmaceutical Research (12:23) Use of AI tools in Medicine (19:56) Biotech Synergy: Collaborative Exploration and Target Discovery (37:17) Decoding Aging with AI (46:52) Aging, Overpopulation, and Drug Breakthroughs (1:07:33) Pharmaceutical Industry Challenges (1:20:42) How AI is Advancing Longevity through Medicine Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

This episode is sponsored by Netsuite by Oracle, the number one cloud financial system, streamlining accounting, financial management, inventory, HR, and more. Download NetSuite's popular KPI Checklist, designed to give you consistently excellent performance - absolutely free, at NetSuite.com/EYEONAI. On episode #133 of the Eye on AI podcast, Craig Smith sits down with Michael Jordan. A revered scientist and distinguished professor at the University of California, Berkeley, Michael's expertise spans machine learning, statistics, and artificial intelligence. In this episode we explore the intricate landscape of deep learning, its statistical bedrock, and the myriad applications of machine learning methods. We dig deeper into the cloud computing revolution, ignited by deep learning, which has empowered behemoths like Amazon to streamline their logistics and commerce data. The dialogue continues as we uncover the trends in deep learning, its role in the grand scheme of AI, and the persisting challenges in the domain. We conclude by contemplating the repercussions of AI and optimization in complex systems. We examine its historical roots in the mid-20th century, its potential to replace jobs, and its application in various sectors such as financial markets, healthcare, and education. (00:00) Preview (01:00) Introduction (01:36) NetSuite by Oracle (03:53) Deep Learning Advancements in AI (12:56) AI Optimization in Complex Systems (28:53) Future of Machine Learning in Healthcare (39:45) Privacy and Value in Multi-Agent Learning (54:48) Learning Systems, Avatars, and Music Business (1:02:20) Single Source of Truth for Business Owners (1:04:28) NetSuite by Oracle Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

On episode #132 of the Eye on AI podcast, Craig Smith sits down with Scott Downes, Chief Technology Officer at Invisible Technologies. We crack open the fascinating world of large language models (LLMs). What are the unique ways LLMs can revolutionize text cleanup, product classification, and more? Scott unpacks the power of technology like Reinforcement Learning for Human Feedback (RLHF) that expands the horizons of data collection. This podcast is a thorough analysis of the world of language and meaning. How does language encode meaning? Can RLHF be the panacea for complex conundrums? Scott breaks down his vision about using RLHF to redefine problem-solving. We dive into the vexing concept of teaching a language model through reinforcement learning without a world model. We discuss the future of the human workforce in AI, hear Scott's insights on the potential shift from labellers to RLHF workers. What implications does this shift hold? Can AI elevate people to work on more complicated tasks? From exploring the economic pressure companies face to the potential for increased productivity from AI, we break down the future of work. (00:00) Preview and introduction (01:33) Generative AI's Dirty Little Secret (17:33) Large Language Models in Problem Solving (23:24) Large Language Models and RLHF Challenges (30:07) Teaching Language Models Through RLHF(35:35) Language Models' Power and Potential(53:00) Future of Human Workforce in AI(1:03:10) AI Changing Your World Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to episode #131 of the Eye on AI podcast with Andrew Ng. Get ready to challenge your perspectives as we sit down with Andrew Ng. We navigate the widely disputed topic of AI as a potential existential threat, with Andrew assuring us that, with time and global cooperation, safety measures can be built to prevent disaster. He offers insight into the debates surrounding the harm AI might cause, including the notions of AI as a bio-weapon and the notorious 'paper clip argument'. Listen as Andrew debunks these theories, delivering an interesting argument for why he believes the associated risks are minimal.Onwards, we venture into the intriguing realm of AI's capability to understand the world, setting the stage for a conversation on how we can objectively assess their comprehension. We explore the safety measures of AI, drawing parallels with the rigour of the aviation industry, and contemplate on the consensus within the research community regarding the danger posed by AI. (00:00) Preview (01:08) Introduction (02:15) Existential risk of artificial intelligence (05:50) Aviation analogy with artificial intelligence (10:00) The threat of AI & deep learning (13:15) Lack of consensus in AI dangers (18:00) How AI can solve climate change (24:00) Landing AI and Andrew Ng (27:30) Visual prompting for images Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI Our sponsor for this episode is Masterworks, an art investing platform. They buy the art outright, from contemporary masters like Picasso and Banksy, then qualify it with the SEC, and offer it as an investment. Net proceeds from its sale are distributed to its investors. Since their inception, they have sold over $45 million dollars worth of artwork And so far, each of Masterworks' exits have returned positive net returns to their investors. Masterworks has over 750,000 users, and their art offerings usually sell out in hours, which is why they've had to make a waitlist. But Eye on AI viewers can skip the line and get priority access right now by clicking this link: https://www.masterworks.art/eyeonai Purchase shares in great masterpieces from artists like Pablo Picasso, Banksy, Andy Warhol, and more. See important Masterworks disclosures: https://www.masterworks.com/cd "Net Return" refers to the annualized internal rate of return net of all fees and costs, calculated from the offering closing date to the date the sale is consummated. IRR may not be indicative of Masterworks paintings not yet sold and past performance is not indicative of future results. Returns shown are 4 examples of midrange returns selected to demonstrate Masterworks performance history. Returns may be higher or lower. Investing involves risk, including loss of principal.

Welcome to episode #130 of Eye on AI with Mathew Lodge. In this episode, we explore the world of reinforcement learning and code generation. Mathew Lodge, the CEO of Diffblue, shares insights into how reinforcement learning fuels generative AI. As we explore the intricacies of reinforcement learning, we uncover its potential in game playing and guiding us towards solutions. We shed light on the products that it powers, such as AlphaGo and AlphaDev. However, we also address the challenges of large language models and explain why they may not be the ultimate solution for code generation. In the last part of our conversation, we delve into the future of language models and intelligence. Mathew shares valuable insights on merging no-code and low-code solutions. We confront the skepticism of software developers towards AI for code products and the task of articulating program outcomes. Wrapping up, we reflect on the evolution of programming languages and the impact of abstraction on machine learning. (00:00) Preview & sponsorship (01:51) Reinforcement Learning and Code Generation (04:39) Reinforcement Learning and Improving Algorithms (15:32) The Challenges of Large Language Models(23:58) Future of Language Models and Intelligence (35:50) Challenges and Potential of AI-generated Code (48:32) Programming Language Evolution and Higher-Level Languages Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to episode #129 of Eye on AI with Alexandra Geese. Navigating the complex waters of the European Union's AI Act is no simple task. Yet, that's exactly what Alexandra Geese, a member of the European Parliament, and I venture to do in this conversation. Alexandra's insights into the AI Act, its four defined categories of AI applications, and its current negotiation phase with the European Council and the European Commission are illuminating. We delve into the Act's essential mission: ensuring AI serves humanity, while also exploring the influence of powerful players in the AI industry on the EU's legislation. As our journey deepens, we tackle a range of crucial issues underpinning the Act. Alexandra and navigate through potential economic implications for those who rely on copyright legislation, and the risk of Europe falling behind in AI implementation if the Act is too restrictive. We also touch on the involvement of major American AI firms in the Act's finalization process, and implications for copyrighted material. We dive into the ongoing debates shaping the legislation and the enforcement of the law, once passed. Alexandra shares her thoughts on potential fines for violations, different AI zones, and the possibility of the US following Europe's lead in AI legislation. We wrap up with a deep reflection on the environmental impact of AI, the power held by few companies, and our collective responsibility as AI reshapes the world. (00:00) Preview (00:52) Introduction (03:10) Alexandra Geese background in digital legislation (04:00) The AI act: the explanation and details (08:00) The foundations of corporations for AI regulation (13:00) Copyright regulation and impacts of creativity (17:00) We need AI that serves humanity (21:30) Are foundation models high risk to society? (25:00) Should people be worried about investing in AI? (30:45) What is dynamic AI regulation? (36:10) What is the timeline for AI regulation? (38:50) What penalties will be applied to AI regulation? (44:30) Will US & EU merge on AI regulation? (50:30) How to solve AI hallucinations Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI Found is a show about founders and company-building that features the change-makers and innovators who are actually doing the work. Each week, TechCrunch Plus reporters, Becca Szkutak and Dom-Madori Davis talk with a founder about what it's really like to build and run a company—from ideation to launch. They talk to founders across many industries and their conversations often lead back to AI as many startups start to implement AI into what they do. New episodes of Found are published every Tuesday and you can find them wherever you listen to podcasts. Found podcast: https://podlink.com/found

Yoshua Bengio, the legendary AI expert, will join us for Episode 128 of Eye on AI podcast. In this episode, we delve into the unnerving question: Could the rise of a superhuman AI signal the downfall of humanity as we know it? Join us as we embark on an exploration of the existential threat posed by superhuman AI, leaving no stone unturned. We dissect the Future of Life Institute's role in overseeing large language model development. As well as the sobering warnings issued by the Centre for AI Safety regarding artificial general intelligence. The stakes have never been higher, and we uncover the pressing need for action. Prepare to confront the disconcerting notion of society's gradual disempowerment and an ever-increasing dependency on AI. We shed light on the challenges of extricating ourselves from this intricate web, where pulling the plug on AI seems almost impossible. Brace yourself for a thought-provoking discussion on the potential psychological effects of realizing that our relentless pursuit of AI advancement may inadvertently jeopardize humanity itself. In this episode, we dare to imagine a future where deep learning amplifies system-2 capabilities, forcing us to develop countermeasures and regulations to mitigate associated risks. We grapple with the possibility of leveraging AI to combat climate change, while treading carefully to prevent catastrophic outcomes. But that's not all. We confront the notion of AI systems acting autonomously, highlighting the critical importance of stringent regulation surrounding their access and usage. (00:00) Preview (00:42) Introduction (03:30) Yoshua Bengio's essay on AI extinction (09:45) Use cases for dangerous uses of AI (12:00) Why are AI risks only happening now? (17:50) Extinction threat and fear with AI & climate change (21:10) Super intelligence and the concerns for humanity (15:02) Yoshua Bengio research in AI safety (29:50) Are corporations a form of artificial intelligence? (31:15) Extinction scenarios by Yoshua Bengio (37:00) AI agency and AI regulation (40:15) Who controls AI for the general public? (45:11) The AI debate in the world Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to Episode 127 of the Eye on AI podcast with host Craig Smith and guest Clemens Mewald. In this episode, we dive into the world of AI and its transformative impact on industries. Join us as we explore Instabase, the cutting-edge company led by our guest, a seasoned engineer with an impressive background at Google Brain, Databricks, and now Instabase. Discover how Instabase is revolutionizing automation and content capture across various organizations using AI-driven methods. Uncover the mission behind Instabase and delve into the intricate details of AI Hub, a groundbreaking marketplace for AI models and products. We explore the limitations of model repositories and marketplaces, particularly in large-scale applications. As we compare AI Hub with the AWS Marketplace, we touch upon the abundance of low-code app development solutions in the market, highlighting Accio's rich SaaS offerings and generative AI apps as an industry benchmark. No discussion about AI would be complete without delving into the potential of GPT-4, a powerful language model capable of accurately predicting task outcomes. Join us on this ride as we uncover the heart of AI, its revolutionary applications, and its transformative power across industries. (00:00) Preview (00:38) Introduction (01:22) Clemens background and Google Brain (02:44) Instabase and solving unstructured data problems (07:40) How Instabase works and different use cases (13:20) The long term vision of the AI Hub (17:12) Blockchain based marketplace for AI models (21:50) AWS Marketplace compared to Instabase (24:05) Generative AI and no code web apps (31:05) Biggest concerns of using Open AI for security (35:40) Considerations of use cases of GPT4 (40:00) LLMs acting as knowledge and reasoning engines (46:40) Using different AI models based on different tasks (51:00) Leveraging other AI models for compatibility (54:14) How to get people to start using Instabase Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to episode #126 of Eye on AI with Craig Smith and Noam Chomsky. Are neural nets the key to understanding the human brain and language acquisition? In this conversation with renowned linguist and cognitive scientist Noam Chomsky, we delve into the limitations of large language models and the ongoing quest to uncover the mysteries of the human mind. Together, we explore the historical development of research in this field, from Minsky's thesis to Jeff Hinton's goals for understanding the brain. We also discuss the potential harms and benefits of large language models, comparing them to the internal combustion engine and its differences from a gazelle running. We tackle the difficult task of studying the neurophysiology of human cognition and the ethical implications of invasive experiments. As we consider language as a natural object, we discuss the works of notable figures such as Albert Einstein, Galileo, Leibniz, and Turing, and the similarities between language and biology. We even entertain the possibility of extraterrestrial language and communication. Join us on this thought-provoking journey as we explore the intricacies of language, the brain, and our place in the cosmos. (00:00) Preview (00:43) Introduction (01:54) Noam Chomsky's neural net ideology & criticisms (6:58) Jeff Hinton & Noam Chomsky's: How the brain works (10:05) Correlation between neural nets and the brain (11:11) Noam Chomsky's reaction to Chat-GPT & LLMs (15:21) Exploring the mechanisms of the brain (19:00) What do we learn from chatbots? (22:30) What are impossible languages? (26:45) Generative AI doesn't show true intelligence? (28:40) Is there a danger of AI becoming too intelligent? (31:30) Can AI language models become sentient? (36:40) Turing machine and neural nets experimentations (42:40) Non-evasive procedures for understanding the brain (45:54) Does Noam Chomsky still work on understanding the brain? (49:33) Is Noam Chomsky excited about the future of neural nets? (55:30) Albert Einstein and Galileo's principles (55:40) Is there an extraterrestrial language model? Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to episode 125 of Eye on AI, where we embark on a journey into the realm of Generative AI. In this episode, we have the pleasure of chatting with Pascal Weinberger, co-founder and CEO of Bardeen AI, who takes us through the evolution of AI and its incredible potential for creativity and professional endeavors. Join us as we venture behind the scenes of Telefonica's Moonshot Lab, where AI projects in healthcare, energy, and city planning are explored. Discover the fascinating ideas and initiatives that have emerged, including the birth of a mental health company, as we uncover the immense impact of Generative AI. During our conversation, we'll delve into the nuances of Generative AI technology, exploring how industry giants like Microsoft and Google are harnessing its power to enhance their products. We'll also discuss the strategies and challenges faced by companies in the competitive Generative AI market, with a strong focus on meeting the needs of end users. We'll also tackle the ongoing debates surrounding the risks and benefits of AI technology, ensuring you stay ahead of the curve in this ever-evolving world of Generative AI. Tune in and join us as we unravel the secrets of Generative AI, paving the way for a future where creativity and productivity reach new heights. (00:00) Preview (00:24) Pascal's Weinberger background in Telefonica (08:28) Machine learning & AI with Pascal's Weinberger (10:28) How Pascal's Weinberger founded Bardeen AI (13:25) Generate AI MVP for Bardeen AI (17:21) Generative AI applications and OpenAI competition (22:24) Competition in the AI space (25:24) Big tech companies vs. startups in AI (31:46) The future of AI and transformer algorithm (32:41) Bardeen AI features and functionality (46:24) AutoGPT problems and considerations (50:54) Risk of AI & misuse of commands Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to episode #124 of the Eye on AI podcast, where we bring you the latest insights into the fascinating world of artificial intelligence. In this episode, Craig Smith is joined by Sina Kian, General Counsel and COO at Aleo, as they dive deep into the revolutionary realm of zero-knowledge proofs. Join us as we explore the incredible potential of zero-knowledge proofs in safeguarding sensitive data while leveraging it for machine learning and AI applications. Sina Kian provides shares how this innovative technology can reshape privacy, digital identity, and even social media authentication. During this conversation, we delve into the power of privacy-preserving blockchain technology and its far-reaching impact across industries. Discover how Aleo is at the forefront of making digital identity more secure and how it can be seamlessly integrated across platforms without compromising sensitive information. We examine the future of machine learning and AI, unraveling the role that digital identity plays in accessing products and content based on location. As we venture into the depths of the social media landscape, we also explore the risks and rewards associated with user data and privacy. Gain insights into how privacy-preserving technology can shield user information and authenticate data and content without compromising privacy. This conversation will discusses the potential of zero-knowledge proofs and privacy-preserving technology, offering a glimpse into how they will shape the future of machine learning and AI. (00:00) Preview (00:41) Introduction (02:28) Sina Kian's background in Aleo & blockchain (05:49) Blockchain's integration with AI & machine learning (11:48) How data is protected in blockchain technology (12:25) Use cases of encryption with Aleo (18:53) How Aleo works with an open source protocol (24:13) Aleo's progress in developing its open source project (31:13) Why social media platforms capture your data (34:16) How can you find widespread adoption? (35:43) How the government are getting involved in digital identity (41:53) How data privacy integrates to Web 3.0 (45:15) Blockchain's implementation in the real world (48:43) Next steps from Aleo (53:53) Social media interaction with privacy Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to Eye on AI, the podcast that keeps you informed about the latest trends, obstacles, and possibilities in the realm of artificial intelligence. In this episode, we have the privilege of engaging in a thought-provoking discussion with Aidan Gomez, an exceptional AI developer and co-founder of Cohere. Aidan's passion lies in enhancing the efficiency of massive neural networks and effectively deploying them in the real world. Drawing from his vast experience, which includes leading a team of researchers at For.ai and conducting groundbreaking research at Google Brain, Aidan provides us with unique insights and anecdotes that shed light on the AI landscape. During our conversation, Aidan explains his collaboration with the legendary Geoffrey Hinton and their remarkable project at Google Brain. We delve into the intricate architecture of AI systems, demystifying the construction of the transformative transformer algorithm. Aidan generously shares his knowledge on the creation of attention within these models and the complexities of scaling such systems. As we explore the fascinating domain of language models, Aidan discusses their learning process, bridging the gap between code and data. We uncover the immense potential of these models to suggest other large-scale counterparts. We gain invaluable insights into Aidan's journey as a co-founder of Cohere, an innovative platform revolutionizing the utilization of language technology. Tune in to Eye on AI now to immerse yourself in a captivating conversation that will expand your understanding of this ever-develop field. (00:00) Preview (00:33) Introduction & sponsorship (02:00) Aidan's background with machine learning & AI (05:10) Geoffrey Hinton & Aidan Gomez working together (07:55) Aidan Gomez & Google Brain's project (12:53) Aidan's role in building AI architecture (15:25) How the transformer algorithm is built (18:25) How do you create attention? (20:40) How do you scale the model? (25:10) How language models learn from code and data (29:55) Did you know the potential of the project? (34:15) Can LLMs suggest other large models? (36:45) How Aidan Gomez started Cohere (41:10) How do people use Cohere? (46:50) Examples of language technology models (48:40) How Cohere handles hallucinations (52:53) The dangers of AI Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to Eye on AI, the podcast that explores the latest developments, challenges, and opportunities in the world of artificial intelligence. In this episode, we sit down with Connor Leahy, an AI researcher and co-founder of EleutherAI, to discuss the darker side of AI. Connor shares his insights on the current negative trajectory of AI, the challenges of keeping superintelligence in a sandbox, and the potential negative implications of large language models such as GPT4. He also discusses the problem of releasing AI to the public and the need for regulatory intervention to ensure alignment with human values. Throughout the podcast, Connor highlights the work of Conjecture, a project focused on advancing alignment in AI, and shares his perspectives on the stages of research and development of this critical issue. If you're interested in understanding the ethical and social implications of AI and the efforts to ensure alignment with human values, this podcast is for you. So join us as we delve into the darker side of AI with Connor Leahy on Eye on AI. (00:00) Preview (00:48) Connor Leahy's background with EleutherAI & Conjecture (03:05) Large language models applications with EleutherAI (06:51) The current negative trajectory of AI (08:46) How difficult is keeping super intelligence in a sandbox? (12:35) How AutoGPT uses ChatGPT to run autonomously (15:15) How GPT4 can be used out of context & negatively (19:30) How OpenAI gives access to nefarious activities (26:39) The problem with the race for AGI (28:51) The goal of Conjecture and advancing alignment (31:04) The problem with releasing AI to the public (33:35) FTC complaint & government intervention in AI (38:13) Technical implementation to fix the alignment issue (44:34) How CoEm is fixing the alignment issue (53:30) Stages of research and development of Conjecture Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this podcast, we sit down with Danny Tobey, an attorney with the global law firm DLA Piper, to discuss the changing legal dynamics surrounding artificial intelligence. As one of the leading experts in the field, Danny provides valuable insights into the current state of legislation and regulation, the efforts of regulatory bodies like the Federal Trade Commission in tackling issues related to AI, and how the law firm of the future will look as AI continues to transform the economy. With the growing impact of AI on all aspects of our lives, the legal profession is facing unique challenges and opportunities. Danny brings a wealth of knowledge and experience to the conversation, having worked with clients in industries ranging from healthcare to financial services to consumer products. Throughout the podcast, Danny explores the ethical and legal implications of AI, as well as the ways in which AI is already reshaping the legal industry. He provides thoughtful perspectives on how the legal profession can adapt and evolve to meet the demands of an AI-driven economy, and the role that lawyers and regulatory bodies will play in shaping the future of this transformative technology. Whether you're a legal professional looking to stay on top of the latest developments in AI, or simply interested in the ways that AI is changing the legal landscape, this podcast is sure to offer valuable insights and food for thought. So join us as we dive deep into the intersection of law and artificial intelligence with Danny Tobey. Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this episode of the Eye on A.I. podcast, host Craig Smith interviews Yoshua Bengio, one of the founding fathers of deep learning and a Turing Award winner. Bengio shares his insights on the famous pause letter, which he signed along with other prominent A.I. researchers, calling for a more responsible approach to the development of A.I. technologies. He discusses the potential risks associated with increasingly powerful A.I. models and the importance of ensuring that models are developed in a way that aligns with our ethical values. Bengio also talks about his latest research on world models and inference machines, which aim to provide A.I. systems with the ability to reason for reality and make more informed decisions. He explains how these models are built and how they could be used in a variety of applications, such as autonomous vehicles and robotics. Throughout the podcast, Bengio emphasises the need for interdisciplinary collaboration and the importance of addressing the ethical implications of A.I. technologies. Don't miss this insightful conversation with one of the most influential figures in A.I. on Eye on A.I. podcast! Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

Welcome to the latest episode of our podcast featuring Edo Liberty, an AI expert and former creator of SageMaker at Amazon's AI labs. In this episode, Edo discusses how his team at Pinecone.io is tackling the problem of hallucinations in large language models like ChatGPT. Edo's approach involves using vector embeddings to create a long-term memory database for large language models. By converting authoritative and trusted information into vectors, and loading them into the database, the system provides a reliable source of information for large language models to draw from, reducing the likelihood of inaccurate responses. Throughout the episode, Edo explains the technical details of his approach and shares some of the potential applications for this technology, including AI systems that rely on language processing. Edo also discusses the future of AI and how this technology could revolutionise the way we interact with computers and machines. With his insights and expertise in the field, this episode is a must-listen for anyone interested in the latest developments in AI and language processing. We have a new sponsor this week: NetSuite by Oracle, a cloud-based enterprise resource planning software to help businesses of any size manage their financials, operations, and customer relationships in a single platform. They've just rolled out a terrific offer: you can defer payments for a full NetSuite implementation for six months. That's no payment and no interest for six months, and you can take advantage of this special financing offer today at netsuite.com/EYEONAI Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this podcast episode, Ilya Sutskever, the co-founder and chief scientist at OpenAI, discusses his vision for the future of artificial intelligence (AI), including large language models like GPT-4. Sutskever starts by explaining the importance of AI research and how OpenAI is working to advance the field. He shares his views on the ethical considerations of AI development and the potential impact of AI on society. The conversation then moves on to large language models and their capabilities. Sutskever talks about the challenges of developing GPT-4 and the limitations of current models. He discusses the potential for large language models to generate a text that is indistinguishable from human writing and how this technology could be used in the future. Sutskever also shares his views on AI-aided democracy and how AI could help solve global problems such as climate change and poverty. He emphasises the importance of building AI systems that are transparent, ethical, and aligned with human values. Throughout the conversation, Sutskever provides insights into the current state of AI research, the challenges facing the field, and his vision for the future of AI. This podcast episode is a must-listen for anyone interested in the intersection of AI, language, and society. Timestamps: (00:04) Introduction of Craig Smith and Ilya Sutskever. (01:00) Sutskever's AI and consciousness interests. (02:30) Sutskever's start in machine learning with Hinton. (03:45) Realization about training large neural networks. (06:33) Convolutional neural network breakthroughs and imagenet. (08:36) Predicting the next thing for unsupervised learning. (10:24) Development of GPT-3 and scaling in deep learning. (11:42) Specific scaling in deep learning and potential discovery. (13:01) Small changes can have big impact. (13:46) Limits of large language models and lack of understanding. (14:32) Difficulty in discussing limits of language models. (15:13) Statistical regularities lead to better understanding of world. (16:33) Limitations of language models and hope for reinforcement learning. (17:52) Teaching neural nets through interaction with humans. (21:44) Multimodal understanding not necessary for language models. (25:28) Autoregressive transformers and high-dimensional distributions. (26:02) Autoregressive transformers work well on images. (27:09) Pixels represented like a string of text. (29:40) Large generative models learn compressed representations of real-world processes. (31:31) Human teachers needed to guide reinforcement learning process. (35:10) Opportunity to teach AI models more skills with less data. (39:57) Desirable to have democratic process for providing information. (41:15) Impossible to understand everything in complicated situations. Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this episode, Ben Sorscher, a PhD student at Stanford, sheds light on the challenges posed by the ever-increasing size of data sets used to train machine learning models, specifically large language models. The sheer size of these data sets has been pushing the limits of scaling, as the cost of training and the environmental impact of the electricity they consume becomes increasingly enormous. As a solution, Ben discusses the concept of "data pruning" - a method of reducing the size of data sets without sacrificing model performance. Data pruning involves selecting the most important or representative data points and removing the rest, resulting in a smaller, more efficient data set that still produces accurate results. Throughout the podcast, Ben delves into the intricacies of data pruning, including the benefits and drawbacks of the technique, the practical considerations for implementing it in machine learning models, and the potential impact it could have on the field of artificial intelligence. Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this episode, Yann LeCun, a renowned computer scientist and AI researcher, shares his insights on the limitations of large language models and how his new joint embedding predictive architecture could help bridge the gap. While large language models have made remarkable strides in natural language processing and understanding, they are still far from perfect. Yann LeCun points out that these models often cannot capture the nuances and complexities of language, leading to inaccuracies and errors. To address this gap, Yann LeCun introduces his new joint embedding predictive architecture - a novel approach to language modelling that combines techniques from computer vision and natural language processing. This approach involves jointly embedding text and images, allowing for more accurate predictions and a better understanding of the relationships between original concepts and objects. Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this episode, Terry Sejnowski, an AI pioneer, chairman of the NeurIPS Foundation, and co-creator of Boltzmann Machines, delves into the latest developments in deep learning and their potential impact on our understanding of the human brain. Terry Sejnowski begins by discussing the NeurIPS conference - one of the most significant events in the field of artificial intelligence - and its role in advancing research and innovation in deep learning. He shares insights into the latest breakthroughs in the field, including the repurposing of the sleep-wake cycle of Boltzmann Machines in Geoff Hinton's new Forward-Forward algorithm. Throughout the episode, Terry Sejnowski shares his expertise on the intersection of artificial intelligence and neuroscience, exploring how advances in deep learning may help us better understand the complexities of the human brain. He discusses how researchers are using AI techniques to study brain activity and the potential implications for fields such as medicine and psychology. Overall, this episode will be of particular interest to those interested in the latest developments in artificial intelligence and their potential applications in neuroscience and related fields. Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

In this episode, Geoffrey Hinton, a renowned computer scientist and a leading expert in deep learning, provides an in-depth exploration of his groundbreaking new learning algorithm - the forward-forward algorithm. Hinton argues this algorithm provides a more plausible model for how the cerebral cortex might learn, and could be the key to unlocking new possibilities in artificial intelligence. Throughout the episode, Hinton discusses the mechanics of the forward-forward algorithm, including how it differs from traditional deep learning models and what makes it more effective. He also provides insights into the potential applications of this new algorithm, such as enabling machines to perform tasks that were previously thought to be exclusive to human cognition. Hinton shares his thoughts on the current state of deep learning and its future prospects, particularly in neuroscience. He explores how advances in deep learning may help us gain a better understanding of our own brains and how we can use this knowledge to create more intelligent machines. Overall, this podcast provides a fascinating glimpse into the latest developments in artificial intelligence and the cutting-edge research being conducted by one of its leading pioneers. Craig Smith Twitter: https://twitter.com/craigssEye on A.I. Twitter: https://twitter.com/EyeOn_AI

To set the stage for some terrific conversations I have coming to you in the new year, in this episode we go back to some earlier conversations that talk about how we got to where we are in deep learning and how those early threads continue to lead innovation.

This week I talk to Bob Rogers, a Harvard trained astrophysicist who once built digital twins of black holes to better understand them, and now builds digital twins of supply chains to help make them more efficient and resilient.

Jonathon Reilly, co-founder of Akkio, a no-code AI platform, talks about how users with a web browser and an idea have the power to bring AI to life themselves without having to write code.

Moses Guttmann, founded ClearML, talks about the evolution of the MLOps industry over the past few years and ClearML's contribution to it.

Bratin Saha, head of Amazon's machine learning services, talks about Amazon's growing dominance in model building and deploying AI, about the company's SageMaker platform, and whether anyone can compete with the behemoth.

Peter Schrammel, one of the founders of Diffblue, an automated unit-test writing software company, speaks about the increasing automatic generation of code and how he sees such automation increasing the productivity of developers.

Michael Kearns, a computer scientist professor at the University of Pennsylvania and an Amazon scholar talks about differential privacy, how Amazon's research approach differs from its peers, and how AI will eventually permeate all aspects of our lives.

Jacki Morie, a senior XPRIZE advisor, talks about the ANA Avatar XPRIZE, a competition focused on creating a physical avatar system that will seamlessly transport human skills and experience to distant locations. The four-year competition is in its final stretch.

Aaron Patzer, founder of the personal finance app MINT and more recently founder of the AI-based healthcare company Vital, talks about keeping customer data private and the promise, giving emergency room patients information with AI and finding friendly solutions to anxiety-producing problems.

Rohit Prasad, Amazon's Senior Vice President and Head Scientist for Alexa, speaks about the development of conversational AI and virtual assistants and the merging of IoT sensor data into ambient intelligence - AI that is always present and immediately accessible.

Ken Washington, who leads Amazon's consumer robotics team, talks about the company's compact wheeled robot called Astro. Ken talked about Astro's evolution, it's popular and possible use cases, and what might be in store in the future.

Almammon Rasool Abdali, a machine-learning engineer and PhD student in Baghdad, is one of the more prominent members of Iraq's small but growing deep learning community. We talked about his work, which involves vision systems to detect violence, and about the state of artificial intelligence research and teaching in the country generally.

Ma Yanjun, General Manager of the AI Technology Ecosystem at Baidu, talks about how Baidu's PaddlePaddle stacks up against other AI frameworks, about Baidu's development of large language models and the direction of AI research in China more generally.

Oriol Vinyals, who leads DeepMind's deep learning team, talks about AlphaCode, his group's code-writing language model, and DeepMind's winding road toward artificial general intelligence.

Tom Denton, a software engineer in Google's bioacoustics group, talks about new algorithms to separate individual bird songs from the cacophony of the forest - and gives some examples. The Eye on AI podcast is sponsored by ClearML, the MLOps solution.

Andrew Ng, founder of Google Brain, Coursera and Landing AI, talks about his vision of data-centric AI, MLOps and the future of supervised vs unsupervised learning. The Eye on AI podcast is sponsored by ClearML.

Tom Siebel, founder and CEO of C3.ai talks about AI projects including military target acquisition and precision healthcare while musing about the dark side of our technological future. The Eye on AI podcast is sponsored by Clear.ML

Max Bileschi, a software engineer at Google Research, talks about his team's application of convolutional neural networks to predict the function of amino acid sequences in a protein. Eye on AI is sponsored by Clear.ML.

Pushmeet Kohli, the head of DeepMind's AI for Science and one of the brains behind AlphaFold, the machine learning system that is helping solve the protein folding problem. The episode is sponsored by Clear.ML, an open-source MLOps solution.

Currently the largest AI system in the world is China's WuDao 2.0, a sparse, multimodal, large language model with 1.75 trillion parameters. Tang Jie, a professor at China's Qinghua University, who leads the WuDao team, talks about how the model was built, why it is unique and what his team plans for the future.

Connor Leahy, one of the minds behind Eleuther AI and its open-source large language model, GPT-J, talks about the building such models and their implications for the future.

Robert O. Work, former Deputy Secretary of Defense and recently co-chairman of the National Security Commission on AI, talks about competition between the US and China to integrate AI into their military capabilities.

Stephen DeAngelis, head of Enterra Solutions, reminds us that so-called Old-Fashioned AI continues to be a powerful tool. He talked about leveraging knowledge bases, inference engines and symbolic logic to make decisions about large dynamic systems.

Daniel Ho, associate director of Stanford's Institute for Human-Centered Artificial Intelligence talks about the proposed National AI Research Resource, an effort to expand the data and compute available to academic researchers, leveling the playing field with researchers in private companies.

Andrew Feldman, one of the founders and CEO of Cerebras Systems, talks about the company's wafer-scale computer chip optimized for machine learning and about the network of chips that company has built that has as much computing power as a human brain.

Adobe's head of research, Gavin Miller, talks about AI-enhanced creativity, guarding against manipulation of visual media and his own AI-enabled robot snakes.

Seth Dobrin, chief AI officer at IBM, talks about the company's tools to increase the trustworthiness, fairness and explainability of AI models.