OpenAI CEO Sam Altman and CTO Mira Murati on the Future of AI and ChatGPT | WSJ Tech Live 2023

2023-10-21に共有
OpenAI's Sam Altman and Mira Murati discuss the capabilities of their future GPT models, how human relationships with AI will change in the future and fears about safety, liability and work as the technology advances.

#OpenAI #ChatGPT #WSJ

コメント (21)
  • @youngforages
    At 83 years old I suddenly have an intense reason to live for another 10 years (and to avoid doubt, it's to watch the developments of this amazing technology).
  • @Paretozen
    In my near 37 years of walking this earth with a curious mind; from exploring the library in my primary school, to receiving my first encyclopedia in hardcover, to the first CD version of Encarta '95, and all the subsequent developments in information access... I can say with full conviction that nothing has revolutionized my thirst for knowledge more than ChatGPT has. It truly is as being reborn, I have no other way to describe it. It is such a pleasure to revisit old topics, and new topics alike, with ChatGPT in hand. Bravo!
  • @HarmonicsAI
    Innovation is easier with small teams making decisive, concentrated bets, who don't tolerate mediocre performers. That's it. — Sam Altman
  • @ankitb3954
    It’s crazy how times change Sam’s out. Feels like the moment when jobs got fired from Apple
  • Sam Altman is friggin awesome! You can tell he processes every question posed to him with a very high degree of calm rationality and the utmost respect to the person who is doing the asking even when the questions themselves are questionable and barbed.
  • @safiya4339
    Am I wrong here? I can easily visualize this scenario; An AI yoga instructor bot, having closely observed say 100 highly rated yoga instructors, and observed their classes and their students' biometrics and progress, takes over as becoming your personal, and personalized, funny, patient, and apparently compassionate yoga teacher. Everyone is saying that "new jobs will be created" but nobody is saying exactly what we will be able to do better than the bots.
  • @reetchow
    The fact that both the audience questions were better than any question the actual interviewer asked is a shame 🤦‍♂
  • @xavier_lucas
    I’m opportuned to have guidance and supervision from some of Wall Street’s eyes and ears. This is an exciting time of year after a tough one all round. I made a significant profit with my MSFT holdings recently and have drawn almost 100k for the holidays.
  • @double051
    Is the audio too soft? I had to turn my phone volume all the way up, and it's not enough
  • @GlennGaasland
    I wish the interviewers for these events would prepare better and ask some new questions, instead of only asking questions that have been asked and answered many times before. This would economize everyone’s time. Here is a great question she could have asked: “Is there a process for creating totally new data, which may be even better than existing data on the internet, to train future models?”
  • The last question @43:15 was the most important - about the need for society (all members of our collective human society) to discuss the oncoming tsunami of extremely rapid job transformations, elimination of industries, needs/costs/timelines for retraining for nearly every field, etc. The question recognizes the challenges & risks for every society, at every worker/employee-level. I found Mira's answer silly. I found Sam's response also ignored the key issue raised -- why aren't AI firms guiding society & govts to compel us to discuss this. AGI and later ASI can fracture parts of our lives. Yet many "brilliant" AI leaders either can't fathom or downplay its many serious risks.
  • @daylanization
    I'm relieved every time I hear an interview with Sam. I'm hoping that many others at OpenAI have a similarly strong moral grounding. I don't mind paying for ChatGPT because OpenAI seems to be taking things in a very responsible direction.
  • I've been following the advancements in AI closely, especially with ChatGPT. It's fascinating to see where this technology is headed.
  • @ColinWPLewis
    Mira consistently cuts through the fog with precise explanations "It is the ultimate technology that can boost creativity and productivity".. Plus as Sam says this is one of the most tremendous technologies that humanity has invented. Thank you for sharing. The hope is that augmented intelligence pushes our civilisation ahead! There are many ways to mess up - but for sure I'm delighted that the OpenAI team are working on the inflection point.
  • Thank you for the upload! After already seeing the juicy excerpts, I can't wait to watch the entire thing.
  • I am very suprised with level and quality of WSJ tech columnist. This interview and questions gives a “school project” vibe. Wasted opportunity to learn more from Sam and Mira.
  • The final question is really great and Mira and Sam's answers are really great. Thank you Mariana Michael(hope I'm getting the name correctly) for bringing the discussion for everyone!
  • I'm curious to hear more on Sam's thoughts on new devices that AGI or GPT can enable.
  • 🎯 Key Takeaways for quick navigation: 00:00 🤖 Introduction and AI's Role in Society - Sam and Mira discuss the role of AI in society. - AI's capacity to excel in specific tasks and its limitations. - The evolving perception of AI's capabilities over time. 02:15 🌐 What is AGI (Artificial General Intelligence)? - Sam explains the concept of AGI (Artificial General Intelligence). - AGI's potential impact on improving the human condition. - The significance of abundant and inexpensive intelligence and energy. 05:54 👤 The Median Human and Human-AI Coexistence - Sam discusses the concept of a "median human" and how AI can assist in tasks where humans are not experts. - The evolving definition of intelligence as AI progresses. - The importance of reliability and safety in AI development. 09:41 🤔 Data and Training AI Models - Sam discusses the use of data in training AI models. - The need for data that people are excited about using. - The evolving role of data as AI models become more capable. 12:39 💰 Economic Arrangements and AI Models - Sam and Mira talk about economic arrangements regarding data. - The importance of creating arrangements that work for everyone. - The shift in focus from memorizing data to reasoning capacity in AI models. 16:52 🗣️ Personalization and Deep Relationships with AI - Discussion of AI's role in building deep relationships and personalization. - The potential for AI systems to be a significant part of people's lives. - The importance of responsible design for AI interactions. 19:38 🕹️ Control and Responsibility in AI Development - Sam and Mira emphasize the shared responsibility in AI development. - The need for a society-wide discussion and decision-making in shaping AI's trajectory. - The recognition that AI development will involve multiple actors and viewpoints. 21:07 🌐 Sam Altman and Mira Murati discuss the future of AI. - Sam and Mira anticipate that the path to AI development will be complex, involving various challenges, competition, societal responses, and regulations. 21:38 🤖 Personalization vs. Human Interaction - Sam expresses his reservations about a future where AI friends are closer to humans than human friends. - They emphasize the importance of distinguishing AI from human interactions by naming it "Chat GPT" and using subtle cues. 22:07 💡 The Coexistence of AI and Human Relationships - The discussion focuses on how people will maintain their human relationships while also interacting with AI, treating them as distinct entities. 23:18 📱 The Ideal Device for AI Interaction - The conversation explores potential devices for interacting with AI, with an emphasis on AI's transformative nature, which may require new computing platforms. 23:48 💼 Exploring New Computing Platforms - Sam discusses the need for a new computing platform that leverages AI's capabilities and how it differs from existing devices like smartphones. 24:26 🧩 Custom Hardware and Chips - The discussion delves into the possibility of OpenAI creating custom hardware or chips and mentions their partnerships with other organizations. 24:54 ⚙️ Alternatives to NVIDIA - Sam talks about alternatives to NVIDIA in the world of AI hardware and mentions their ongoing discussions with ARM. 25:27 💰 Raising Capital - The discussion touches on OpenAI's need for significant capital to achieve its mission, emphasizing transparency in their capital-raising plans. 26:00 📈 Business Growth - Sam and Mira discuss the areas of OpenAI that are currently experiencing the most growth, particularly on the product and revenue sides. 26:58 📣 Subscription Plans and Family Plans - The focus shifts to subscription plans for Chat GPT and the potential for introducing family plans, highlighting the cost and need for flexibility. 27:33 🌐 A.I. Influence and Misinformation - The discussion delves into concerns about AI-generated content, misinformation, and the challenges of addressing individualized persuasion. 28:37 🔐 Responsibility and Regulation - Sam and Mira discuss the role of regulation and collective decision-making in shaping the development and deployment of AI technologies. 29:37 🌐 Adapting to New Technologies - The conversation highlights the need for adaptation and societal infrastructure to support the responsible deployment of AI capabilities. 30:09 🔍 Detecting AI-Generated Content - Sam and Mira discuss OpenAI's efforts to detect AI-generated content in images and text and the need for a nuanced approach. 30:36 💡 User Engagement and Feedback - The conversation emphasizes the importance of user feedback and controlled deployment when addressing the challenges of AI-generated content. 31:31 💧 Challenges in Misinformation Mitigation - The discussion explores the complexity of misinformation issues, requiring interaction and input from various stakeholders. 31:58 💧 Adapting to Reality and User Impact - The conversation highlights the need to roll out AI technologies while making contact with reality and adapting to various implications. 32:56 💡 The Role of Self-Regulation - Sam explains how AI's deployment in the real world and societal adaptation contribute to safety and responsible use, reducing the need for excessive self-regulation. 33:55 💬 Nuances of AI-Generated Content - The discussion highlights the challenges and nuances in addressing AI-generated content, balancing user flexibility and societal impact. 35:24 🌐 Regulation for Advanced AI Models - Sam emphasizes the need for international regulation for the most powerful AI models as they approach superintelligence levels. 36:23 📜 The Role of Watermarking in Content - The discussion touches on the complex issue of watermarking AI-generated content, considering varying opinions and potential consequences. 37:53 📸 Detection Technology for AI-Generated Content - Sam and Mira discuss their efforts to develop detection technology for AI-generated content, both in images and text. 38:52 🔍 User Preferences and Watermarking - The conversation acknowledges differing preferences among users regarding watermarking AI-generated content and how this reflects the evolving landscape. 39:53 🤔 Hopes and Fears for the Future - Sam and Mira share their hope for a brighter future with AI while addressing their concerns, particularly regarding AI's safety and responsible development. 41:01 🚀 Sam Altman's Bunker - A light-hearted reference to Sam's preparedness for an uncertain future, alluding to the idea of a bunker, followed by audience questions. 41:33 🤔 Hopes and Fears for the Future - Sam and Mira discuss their hopes and fears for the future of AI and society. - They emphasize the opportunity to enhance civilization through AI while acknowledging the risks and challenges involved. 43:16 📚 The Continuous Evolution of AI - Sam discusses the continuous nature of AI development, explaining that the path to AGI is an ongoing process rather than a specific inflection point. - He compares AI's development to a constantly evolving exponential curve. 44:20 🌐 Preparing for the Changing Nature of Work - A question is raised about the significant impact of AI on the nature of work and the potential societal divisions it may cause. - Sam and Mira stress the importance of making AI accessible to the wider population and preparing institutions for the workforce changes. 47:20 💼 Adapting to Job Market Changes - Sam discusses the historical context of job market changes due to technological revolutions and their adaptation over generations. - He emphasizes the need for society to address the speed at which AI-driven job market changes will occur and provide people with agency in shaping the future. 48:51 🌐 The Human Desire for Creativity and Contribution - Sam discusses the innate human desire for creative expression and contributing to society, asserting that this aspect of humanity will persist. - He highlights the need for people to actively influence the future rather than relying solely on univer
  • @KrisTC
    Thanks for the family plan question! I definitely want that