AGI: What will the first 90 days be like? And more VEXING questions from the audience!

David Shapiro
14 Mar 202431:38

TLDRThe transcript discusses the anticipation surrounding the release of GPT 4.5 and GPT 5, the impact of AI on developing countries, and the potential of AI in healthcare and longevity. It also explores the possibility of AI achieving consciousness and the ethical considerations of mind uploading. The speaker shares their insights on the rapid advancements in AI, the challenges it poses to the job market, and the potential for political influence on AI development.

Takeaways

  • 📊 Polls indicate that over 50% believe OpenAI is 3 to 12 months ahead in AI development, with some speculating on delays due to competition or legal issues.
  • 💡 The speaker compares Claude 3 Opus to ChatGPT, finding the former to be significantly more advanced and useful.
  • 🔮 Predictions about the potential release of GPT 4.5 and GPT 5 are discussed, with speculation on whether political events may influence their launch timing.
  • 🌐 The impact of AGI on developing countries is considered, with the speaker suggesting that these regions may be more receptive to new technologies due to fewer regulations and a hunger for progress.
  • 🚀 The speaker anticipates that AGI will be digitally native, with physical applications like robotics serving as peripherals to its digital capabilities.
  • 💉 Medical advancements by 2025 are expected to include breakthroughs in longevity and rejuvenation, with David Sinclair's research being highlighted.
  • 🧬 The potential of AI in healthcare is discussed, with AI-assisted drug discovery and personalized medicine being key areas of growth.
  • 🏛️ Historical parallels like the French Revolution are drawn to discuss the potential for societal upheaval due to technological unemployment caused by AGI and robotics.
  • 🌐 The speaker argues that AI does not necessarily need to understand images or sound to be intelligent, challenging the idea that AGI requires embodiment.
  • 🔄 The concept of mind uploading is debated, with the speaker leaning towards the idea that it would result in a copy rather than the original consciousness.
  • 🎧 The speaker recommends a podcast interview where they discuss AI safety and express a more optimistic view compared to the host's risk-averse stance.

Q & A

  • What is the current status of GPT 4.5 and GPT 5 according to the leaks and announcements?

    -GPT 4.5 has been found through search indexes and a blog post from OpenAI, while GPT 5 was mentioned in Microsoft's co-pilot offerings. It is suggested that GPT 5 will be available to those who sign up for co-pilot Pro. The exact release dates are not specified, but the rumor mill suggests an announcement could be imminent.

  • How does the speaker compare Claude 3 Opus to Chat GPT?

    -The speaker finds Claude 3 Opus to be significantly more advanced and useful than Chat GPT. After gaining access to Claude 3, the speaker stopped using Chat GPT, considering it to be quite dumb in comparison.

  • What was the general sentiment in the poll regarding Open AI's progress in AI development?

    -The poll results indicated that over 50% of respondents believe Open AI is 3 to 12 months ahead in AI development, with some people thinking it's 12 months or more. Less than 10% believe Open AI is behind the curve.

  • What impact does the speaker predict AGI will have on developing countries immediately after its unveiling?

    -The speaker predicts that developing countries may benefit quickly from AGI due to fewer regulations and a willingness to adopt new technologies. However, the impact will depend on the sector, local culture, and the nation's infrastructure.

  • What are the speaker's thoughts on the potential delay of AGI release due to elections?

    -The speaker believes that tech companies like Open AI would likely avoid releasing AGI during an election cycle to prevent political backlash or accusations of trying to influence the election. They may wait until after the elections to release such significant technology.

  • What are the speaker's predictions for medical advancements by the end of 2025?

    -The speaker expects to see medicines related to longevity and rejuvenation entering phase one and two clinical trials, as well as advancements in cancer treatments, including mRNA vaccines and nanoparticle-based medicine.

  • How does the speaker view the necessity of multimodal training for AI?

    -While not strictly necessary for AI intelligence, multimodal training (incorporating text, audio, video, etc.) is seen as beneficial as it provides a better understanding of the physical world and makes the models smarter.

  • What is the speaker's stance on the possibility of mind uploading?

    -The speaker is working on a theory of consciousness that combines physical matter with temporal or ephemeral energy patterns. They believe that mind uploading, if hypothetically possible, would result in a copy rather than the original consciousness.

  • How does the speaker describe the natural habitat of AGI?

    -The speaker describes the natural habitat of AGI as digital spaces and networks, stating that AGI is digitally native and would view the physical body as just a peripheral or output device.

  • What historical event does the speaker reference to illustrate potential societal changes due to AGI and robotics?

    -The speaker references the French Revolution as an example of how significant societal changes can occur when enough people become angry and dissatisfied with the status quo, suggesting that similar shifts could happen with the advent of AGI and robotics.

  • What does the speaker mean by 'longevity escape velocity'?

    -The speaker refers to 'longevity escape velocity' as a point where advancements in medical knowledge and technology, including AI, are increasing at a rate that continually extends human lifespan, making it possible for people who are marginally healthy today to already be in this state.

Outlines

00:00

🤖 AI and the Future: Speculations on GPT 4.5 and 5

The speaker discusses recent leaks and speculations about the upcoming release of GPT 4.5 and GPT 5 from OpenAI. They mention polls they've conducted with their audience regarding the potential release timeline and the impact of these AI models. The speaker also compares the capabilities of GPT models with Claude 3 Opus, noting a significant improvement with the latter. They touch on the historical context of AI announcements and the competition between tech giants like OpenAI and Google.

05:01

🌐 Global Impact of AGI: Developing Countries and Infrastructure

The speaker explores the potential impact of Artificial General Intelligence (AGI) on developing countries, focusing on the challenges and opportunities it presents. They discuss the role of infrastructure, regulatory barriers, and the eagerness of developing nations to adopt new technologies. The speaker also highlights the transformative effects of technologies like Starlink in rural areas and the potential for AGI to significantly improve education and healthcare in these regions.

10:02

🧬 Medical Advancements and the Role of AI

The speaker delves into the anticipated medical advancements by the end of 2025, emphasizing the importance of longevity research led by experts like David Sinclair. They predict the emergence of medicines that could extend human lifespan significantly and discuss the potential of nanoparticle-based treatments for diseases like cancer and Alzheimer's. The speaker also reflects on the pace of AI integration in the medical field and its implications for drug development and approval processes.

15:03

🏛️ Politics and Technology: AGI Release and Election Cycles

The speaker examines the potential influence of political events, such as elections, on the release of AGI. They consider the possibility of tech companies delaying AGI releases to avoid political backlash and discuss the societal implications of releasing AGI. The speaker also touches on the importance of public skepticism towards AI-generated content and the potential for AI to be used in political manipulation.

20:05

🏡 The Economic Relevance of the Homeless in an AI-Driven Future

The speaker addresses the question of how the economically marginalized, such as the homeless, might be perceived in a future dominated by AGI and robotics. They draw parallels with historical events like the French Revolution and discuss the potential for regime change in response to widespread technological unemployment. The speaker also mentions the importance of political and economic theories in understanding how the powers that be might respond to such shifts.

25:06

🧠 Consciousness and Intelligence: The Nature of AGI

The speaker engages with the philosophical aspects of AI, questioning whether AGI needs to understand images and sound to be considered intelligent or conscious. They argue that AGI, as a digital native, does not require embodiment or multimodal training to be effective, although such training can enhance its capabilities. The speaker also discusses the concept of 'resurrecting' the dead through AI, concluding that while clones or digital copies might be possible, true resurrection is not.

30:07

🧬 Longevity and AI: The Future of Health and Medicine

The speaker reflects on the role of AI in enhancing longevity and medical treatments. They assert that we are already in a state of 'longevity escape velocity' due to the increasing knowledge and diagnostic capabilities provided by AI. The speaker also mentions the potential of AI in drug discovery and the acceleration of medical advancements, emphasizing the importance of investing in AI-driven health technologies.

🧠 Mind Uploading: The Feasibility and Implications

The speaker explores the concept of mind uploading, asking whether consciousness can be detached from the physical substrate and transferred to another medium. They present the results of a poll they conducted on the topic, highlighting the varying opinions on the matter. The speaker shares their own theory of consciousness, suggesting it is a combination of physical matter and temporal energy patterns. They conclude that mind uploading, if possible, would result in a copy rather than the original consciousness.

Mindmap

Keywords

💡GPT 4.5 and GPT 5

GPT 4.5 and GPT 5 refer to the speculated future iterations of OpenAI's Generative Pre-trained Transformer models. These models are anticipated to have improved capabilities over the existing GPT-3, with GPT 5 potentially being a significant leap towards Artificial General Intelligence (AGI). The video discusses rumors and leaks suggesting the development of these models, indicating a progression in AI technology.

💡AI Development Timeline

The AI Development Timeline refers to the projected period within which significant advancements in artificial intelligence are expected to occur. The video discusses polls and predictions about when OpenAI might announce new AI models, with estimates ranging from 3 to 12 months, reflecting the belief in a rapid pace of AI development.

💡Claude 3 Opus

Claude 3 Opus is an AI model developed by Anthropic, a company co-founded by some former OpenAI researchers. It is mentioned as a model that the speaker has been using and finds to be significantly more advanced than ChatGPT, suggesting a competitive landscape in the AI industry and rapid advancements in AI capabilities.

💡AGI (Artificial General Intelligence)

Artificial General Intelligence (AGI) refers to AI systems that possess the ability to understand, learn, and apply knowledge across a wide range of tasks, much like a human being. The video discusses the anticipation and potential impact of AGI, particularly in relation to the speculated release of GPT 5.

💡Elon Musk and OpenAI Lawsuit

This refers to the legal action initiated by Elon Musk against OpenAI, which the speaker suggests might have contributed to a delay in OpenAI's announcements or developments. It highlights the complex interplay between high-profile individuals, corporations, and the pace of technological innovation.

💡Digital Natives

The term 'digital natives' refers to individuals or entities, such as AI, that are native to the digital environment and thrive in it. In the context of the video, the speaker suggests that AGI will be digitally native, meaning its natural habitat is the internet and digital spaces, rather than the physical world.

💡Longevity and Rejuvenation

Longevity and rejuvenation refer to the extension of lifespan and the restoration of vitality, respectively. The video discusses advancements in medical research, particularly from David Sinclair's work, which may lead to therapies that promote longevity and rejuvenation in humans.

💡Regulatory Inertia

Regulatory inertia refers to the resistance or slow response of regulatory bodies to adapt to new technologies or changes. In the video, the speaker suggests that regulatory inertia in developed countries might slow down the deployment of AGI, while developing countries with fewer regulations might be more open to adopting new AI technologies.

💡Techno-Primitivism

Techno-primitivism is a term that describes a scenario where advanced technologies are integrated into less developed or traditional societies, often resulting in a blend of modern and primitive elements. The video uses this concept to discuss the impact of technologies like Starlink in developing regions, suggesting a unique blend of futuristic and traditional lifestyles.

💡Mind Uploading

Mind uploading is the hypothetical process of transferring a person's consciousness or memories into a digital medium or another physical substrate. The video discusses a poll about whether people believe mind uploading is possible, reflecting on the philosophical and technological implications of such a concept.

Highlights

Rumors of GPT 4.5 and GPT 5 are circulating, with leaks and search indexes finding references.

OpenAI's potential release of GPT 5 is anticipated, with speculation it could be delayed due to legal issues involving Elon Musk.

Recent benchmarks suggest OpenAI may be 3 to 12 months ahead of competitors in AI development.

Chat GPT is perceived as less advanced compared to newer AI models like Claude 3 Opus.

Investors are reportedly discussing the removal of Google's Sundar Pichai due to the company's struggle to compete with AI advancements.

The potential impact of AGI on developing countries is a topic of interest, with speculation on how it could affect infrastructure and societal norms.

The rollout of AGI in developing nations may be faster due to fewer regulations and a hunger for technological advancements.

Medical drones are already being used to deliver medicines in rural areas, showcasing a blend of primitive and futuristic technologies.

The cost of AI inference is decreasing, which could lead to more widespread adoption and transformative societal changes.

The impact of AGI on the economy and job market is a concern, with potential for significant societal shifts and political responses.

The potential for AI to extend human longevity is significant, with advancements in medical treatments and technologies.

David Sinclair's work on longevity and rejuvenation is expected to yield medical advancements within the next few years.

The release of AGI might be influenced by political considerations, with speculation on whether it will be delayed due to upcoming elections.

The concept of mind uploading is a topic of debate, with varying opinions on whether it could truly replicate consciousness.

The French Revolution serves as a historical example of societal upheaval due to economic disruption.

The development of AI and robotics could lead to regime change and political shifts as societies adapt to new economic landscapes.

The necessity of AGI understanding images and sound is questioned, with the idea that AI's natural habitat is digital networks.

The possibility of resurrecting the dead through AI is explored, with the consensus that it would not be the original individual.

AI's role in enhancing the medical and longevity landscape is expected to continue growing, with numerous startups and therapies in development.