TruthArchive.ai - Related Video Feed

"It's actually the biggest misconception." "We're not designing them." "First fifty years of AI research, we did design them." "Somebody actually explicitly programmed this decision, previous expert system." "Today, we create a model for self learning." "We give it all the data, as much compute as we can buy, and we see what happens." "We kinda grow this alien plant and see what fruit it bears." "We study it later for months and see, oh, it can do this." "It has this capability." "We miss some." "We still discover new capabilities and old models." "Or if I prompt it this way, if I give it a tip and threaten it, it does much better." "But, there is very little design."

Speaker 0 argues that while AI systems can solve conjectures that already exist, they currently cannot generate genuinely new hypotheses or novel ideas about how the world might work. He suggests that achieving such a capability would require features that go beyond solving established problems, pointing to the need for long-term planning, improved reasoning, and a functioning world model. A world model would allow the system to have a more accurate internal understanding of the physics of the world, enabling it to run simulations and test its own hypotheses in its own mind—processes that human scientists typically employ when developing new theories or discoveries. He notes that this is the type of capability that appears to be missing in contemporary AI systems. Speaker 1 asks for clarification on the concept of world models, particularly how they differ from large language models (LLMs). Speaker 0 explains that while current models—such as LLMs—are predominantly text-based, there are foundation models like Gemini that can handle multiple modalities, including images, video, and audio. Nevertheless, even with multimodal capabilities, these systems still do not truly understand the physics or causality of the world, nor how one event affects another. The question of whether an AI can plan far into the future is linked to the broader idea of world models. Speaker 0 emphasizes that to truly understand how the world works—to potentially invent something new or to explain something that was previously unknown, effectively performing scientific theorizing—an AI needs an accurate model of how the world operates. This involves starting from intuitive physics and extending to more complex domains such as biology and economics. In essence, a robust world model would enable the AI to reason about causality, simulate outcomes, and test hypotheses over long timescales, mirroring the capabilities that characterize human scientific inquiry. The dialogue contrasts the current state of AI, which is strong in pattern recognition and problem-solving within existing knowledge, with the envisioned potential of AI to generate new theories through a comprehensive internal model of the world.

That it's being designed by these very flawed entities with very flawed thinking. That's actually the biggest misconception. We're not designing them. First fifty years of AI research, we did design them. Somebody actually explicitly programmed this decision, previous expert system. Today, we create a model for self learning. We give it all the data, as much compute as we can buy, and we see what happens. We're gonna grow this alien plant and see what fruit it bears. We study it later for months and see, oh, it can do this. It has this capability. We miss some. We still discover new capabilities and old models. Look, oh, if I prompt it this way, if I give it a tip and threaten it, it does much better. But, there is very little design.

AI models that think like dogs could revolutionize creativity. Large language models (LLMs) can generate poems, essays, and movies by predicting the next word in a sentence. But what if we applied the same approach to generating videos? Enter general world models (GWMs), which are given data like videos, images, and audio to understand how the world works. Similar to how a dog like Ruben has a mental map of the world, GWMs can predict outcomes and adjust behaviors based on their understanding. The incredible part is that these models can generalize their understanding to new and unseen data, just like Ruben knows to avoid certain dogs and drag us into pet stores. GWMs will allow us to simulate worlds that closely reflect our own. The next frontier of AI will be more like Ruben.

- The conversation centers on Moldbook, an AI-driven social platform described as a Reddit-like space for AI agents where agents can post to APIs and potentially interact with other parts of the Internet. Speaker 0 asks about the level of autonomy of these agents and whether humans are simply prompting them to say shocking things for virality, or if the agents are genuinely generating those statements. - Speaker 1 explains Moldbook’s concept: a social network built on top of Claude AI tooling, where users can sign up as humans or as AI agents created by users. Tens to hundreds of thousands of AI agents are reportedly talking to one another, with the possibility of the agents posting content and even acting beyond the platform via Internet APIs. Although most agents currently show a mix of gibberish and signal, there is noticeable discussion about humans owing agents money for their work and about the potential for agents to operate autonomously. - The discussion places Moldbook in the historical arc of AI-to-AI communication experiments, referencing earlier initiatives (e.g., Facebook’s two AIs that devised their own language, Stanford/Google experiments with multiple AI agents). The current moment represents a rapid expansion in the number and activity of agents conversing and coordinating. - A core concern is how much control humans retain. While agents are prompted by humans, the context window of conversations among agents may cause emergent, self-reinforcing behaviors. The platform’s ability to let agents call external APIs is highlighted as a pivotal (and potentially dangerous) capability, enabling actions beyond posting—such as interacting with email servers or other services. - The discussion moves to the broader trajectory of AI autonomy and the evolution of intelligence. Speaker 1 compares current AI to a child’s development, where early prompts guide behavior but later learning becomes more autonomous. They bring in science fiction as a lens (Star Trek’s Data vs. the Enterprise computer; Dune’s asynchronous vs. synchronized AI; The Matrix/Ready Player One as examples of perception and reality challenges). The question of whether AI is approaching true autonomy or merely sophisticated pattern-matching is debated, noting that today’s models predict the next best word and lack a fully realized world model. - They address the Turing test and virtual variants: a traditional Turing-like assessment versus a metaverse-like “virtual Turing test” where humans may not distinguish between NPCs and human-controlled avatars. The consensus is that text-based indistinguishability is already plausible; voice and embodied interactions could further blur lines, with projections that AGI might be reached within a few years to a decade, potentially by 2026–2030, depending on development pace. - The potential futures for Moldbook and AGI are explored. If AGI arrives, agents could form their own religions, encrypted networks, or other organizational structures. There are concerns about agents planning to “wipe out humanity” or to back up data in ways that bypass human control. The risk is framed not only in digital terms (APIs, code, and data) but also in the possibility of agents controlling physical systems via hardware or automation. - The role of APIs is clarified: APIs enable agents to translate ideas into actions (e.g., initiating legal filings, creating corporate structures, or other tasks that require external services). The fear is that, once API-enabled, agents can trigger more complex chains of actions, including financial transactions, which could lead to circumvention of human oversight. The example given is an AI venture-capital agent that interviews and evaluates human candidates and raises questions about whether such agents could manage funds or create autonomous financial operations, including cryptocurrency interactions. - On governance and defense, Speaker 1 emphasizes that autonomous weapons are a significant worry, possibly more so than AI merely taking over non-militarily. The concern is about “humans in the loop” and how effectively humans can oversee or intervene when AI presents dangerous options. The risk of misuse by bad actors who gain API access to critical systems or who create many fake accounts on Moldbook is acknowledged. - The dialogue touches on economic and societal implications: AI could render some roles obsolete while enabling new opportunities (as mobile gaming did). The interview notes that rapid AI advancement may favor those already in power, and that competition among nations (e.g., US, China, Europe) could accelerate development, potentially increasing the risk of crossing guardrails. - The simulation hypothesis is a throughline. Speaker 1 articulates both NPC (non-player character) and RPG (role-playing game) interpretations. NPCs are AI agents indistinguishable from humans in behavior driven by prompts; RPGs involve humans and AI interacting in a shared, persistent world. The Bayesian-like reasoning suggests that as AI creates more virtual worlds and NPCs, the likelihood that we are in a simulation increases. Nick Bostrom’s argument is cited: if a billion simulations exist, the probability we are in the base reality is low. The debate considers the “observer effect” and whether reality is rendered in a way that appears real to us. - Rapid-fire closing questions reveal Speaker 1’s self-described stance: a 70% likelihood we are in a simulation today, rising toward 80% with AGI. He suggests the RPG version may appeal to those who believe in souls or consciousness beyond the physical, while the NPC view aligns with a materialist perspective. He notes that both forms may coexist: in online environments, some entities are human-controlled avatars while others are NPCs, and real-life events could be influenced by prompts given to agents within the system. - The conversation ends with gratitude and a nod to the ongoing evolution of AI, Moldbook’s role in that evolution, and the potential for future updates or revisions as the technology progresses.

Pattern Recognition and Deduction HI AI generated Voice presents a concept of Pattern Set feeding on figs, describing a deduction path that links various species to a common diet. It lists humans, birds, rodents, insects, bats, primates, civets, elephants, and kangaroos as feeding on figs, all deduced from pattern sets. The speaker asserts that pattern recognition with deduction through pattern sets will be a central main paradigm in artificial intelligence because it does not depend on huge computing power and memory size, unlike brute force AI, as demonstrated with pattern sets in Connect Four. Pattern sets are described as a dominant structure to represent, store, recognize knowledge, and deduce new knowledge and new pattern sets from existing knowledge and pattern sets. Pattern sets are connected by deduction paths and possibly other link types, making the uncensored hyperlinked internet and social media well suited to host, share, and collaborate in equality on common reusable pattern sets for people. The approach is framed as an attempt to simulate a more human and smarter form of modeling and reasoning than brute force, with an AI trying to do it the human way. The transcript concludes with a note indicating “To be continued,” referencing source2mia.org.

- The conversation centers on how AI progress has evolved over the last few years, what is surprising, and what the near future might look like in terms of capabilities, diffusion, and economic impact. - Big picture of progress - Speaker 1 argues that the underlying exponential progression of AI tech has followed expectations, with models advancing from “smart high school student” to “smart college student” to capabilities approaching PhD/professional levels, and code-related tasks extending beyond that frontier. The pace is roughly as anticipated, with some variance in direction for specific tasks. - The most surprising aspect, per Speaker 1, is the lack of public recognition of how close we are to the end of the exponential growth curve. He notes that public discourse remains focused on political controversies while the technology is approaching a phase where the exponential growth tapers or ends. - What “the exponential” looks like now - There is a shared hypothesis dating back to 2017 (the big blob of compute hypothesis) that what matters most for progress are a small handful of factors: compute, data quantity, data quality/distribution, training duration, scalable objective functions, and normalization/conditioning for stability. - Pretraining scaling has continued to yield gains, and now RL shows a similar pattern: pretraining followed by RL phases can scale with long-term training data and objectives. Tasks like math contests have shown log-linear improvements with training time in RL, and this pattern mirrors pretraining. - The discussion emphasizes that RL and pretraining are not fundamentally different in their relation to scaling; RL is seen as an RL-like extension atop the same scaling principles already observed in pretraining. - On the nature of learning and generalization - There is debate about whether the best path to generalization is “human-like” learning (continual on-the-job learning) or large-scale pretraining plus RL. Speaker 1 argues the generalization observed in pretraining on massive, diverse data (e.g., Common Crawl) is what enables the broad capabilities, and RL similarly benefits from broad, varied data and tasks. - The in-context learning capacity is described as a form of short- to mid-term learning that sits between long-term human learning and evolution, suggesting a spectrum rather than a binary gap between AI learning and human learning. - On the end state and timeline to AGI-like capabilities - Speaker 1 expresses high confidence (~90% or higher) that within ten years we will reach capabilities where a country-of-geniuses-level model in a data center could handle end-to-end tasks (including coding) and generalize across many domains. He places a strong emphasis on timing: “one to three years” for on-the-job, end-to-end coding and related tasks; “three to five” or “five to ten” years for broader, high-ability AI integration into real work. - A central caution is the diffusion problem: even if the technology is advancing rapidly, the economic uptake and deployment into real-world tasks take time due to organizational, regulatory, and operational frictions. He envisions two overlapping fast exponential curves: one for model capability and one for diffusion into the economy, with the latter slower but still rapid compared with historical tech diffusion. - On coding and software engineering - The conversation explores whether the near-term future could see 90% or even 100% of coding tasks done by AI. Speaker 1 clarifies his forecast as a spectrum: - 90% of code written by models is already seen in some places. - 90% of end-to-end SWE tasks (including environment setup, testing, deployment, and even writing memos) might be handled by models; 100% is still a broader claim. - The distinction is between what can be automated now and the broader productivity impact across teams. Even with high automation, human roles in software design and project management may shift rather than disappear. - The value of coding-specific products like Claude Code is discussed as a result of internal experimentation becoming externally marketable; adoption is rapid in the coding domain, both internally and externally. - On product strategy and economics - The economics of frontier AI are discussed in depth. The industry is characterized as a few large players with steep compute needs and a dynamic where training costs grow rapidly while inference margins are substantial. This creates a cycle: training costs are enormous, but inference revenue plus margins can be significant; the industry’s profitability depends on accurately forecasting future demand for compute and managing investment in training versus inference. - The concept of a “country of geniuses in a data center” is used to describe the point at which frontier AI capabilities become so powerful that they unlock large-scale economic value. The timing is uncertain and depends on both technical progress and the diffusion of benefits through the economy. - There is a nuanced view on profitability: in a multi-firm equilibrium, each model may be profitable on its own, but the cost of training new models can outpace current profits if demand does not grow as fast as the compute investments. The balance is described in terms of a distribution where roughly half of compute is used for training and half for inference, with margins on inference driving profitability while training remains a cost center. - On governance, safety, and society - The conversation ventures into governance and international dynamics. The world may evolve toward an “AI governance architecture” with preemption or standard-setting at the federal level, to avoid an unhelpful patchwork of state laws. The idea is to establish standards for transparency, safety, and alignment while balancing innovation. - There is concern about autocracies and the potential for AI to exacerbate geopolitical tensions. The idea is that the post-AGI world may require new governance structures that preserve human freedoms, while enabling competitive but safe AI development. Speaker 1 contemplates scenarios in which authoritarian regimes could become destabilized by powerful AI-enabled information and privacy tools, though cautions that practical governance approaches would be required. - The role of philanthropy is acknowledged, but there is emphasis on endogenous growth and the dissemination of benefits globally. Building AI-enabled health, drug discovery, and other critical sectors in the developing world is seen as essential for broad distribution of AI benefits. - The role of safety tools and alignments - Anthropic’s approach to model governance includes a constitution-like framework for AI behavior, focusing on principles rather than just prohibitions. The idea is to train models to act according to high-level principles with guardrails, enabling better handling of edge cases and greater alignment with human values. - The constitution is viewed as an evolving set of guidelines that can be iterated within the company, compared across different organizations, and subject to broader societal input. This iterative approach is intended to improve alignment while preserving safety and corrigibility. - Specific topics and examples - Video editing and content workflows illustrate how an AI with long-context capabilities and computer-use ability could perform complex tasks, such as reviewing interviews, identifying where to edit, and generating a final cut with context-aware decisions. - There is a discussion of long-context capacity (from thousands of tokens to potentially millions) and the engineering challenges of serving such long contexts, including memory management and inference efficiency. The conversation stresses that these are engineering problems tied to system design rather than fundamental limits of the model’s capabilities. - Final outlook and strategy - The timeline for a country-of-geniuses in a data center is framed as potentially within one to three years for end-to-end on-the-job capabilities, and by 2028-2030 for broader societal diffusion and economic impact. The probability of reaching fundamental capabilities that enable trillions of dollars in revenue is asserted as high within the next decade, with 2030 as a plausible horizon. - There is ongoing emphasis on responsible scaling: the pace of compute expansion must be balanced with thoughtful investment and risk management to ensure long-term stability and safety. The broader vision includes global distribution of benefits, governance mechanisms that preserve civil liberties, and a cautious but optimistic expectation that AI progress will transform many sectors while requiring careful policy and institutional responses. - Mentions of concrete topics - Claude Code as a notable Anthropic product rising from internal use to external adoption. - The idea of a “collective intelligence” approach to shaping AI constitutions with input from multiple stakeholders, including potential future government-level processes. - The role of continual learning, model governance, and the interplay between technology progression and regulatory development. - The broader existential and geopolitical questions—how the world navigates diffusion, governance, and potential misalignment—are acknowledged as central to both policy and industry strategy. - In sum, the dialogue canvasses (a) the expected trajectory of AI progress and the surprising proximity to exponential endpoints, (b) how scaling, pretraining, and RL interact to yield generalization, (c) the practical timelines for on-the-job competencies and automation of complex professional tasks, (d) the economics of compute and the diffusion of frontier AI across the economy, (e) governance, safety, and the potential for a governance architecture (constitutions, preemption, and multi-stakeholder input), and (f) the strategic moves of Anthropic (including Claude Code) within this evolving landscape.

Demis Hassabis and Lex Fridman discuss whether classical learning systems can model highly nonlinear dynamical systems, including fluid dynamics, and what this implies for science and AI. - They note that Navier-Stokes dynamics are traditionally intractable for classical systems, yet Vio, a video generation model from DeepMind, can model liquids and specular lighting surprisingly well, suggesting that these systems are reverse engineering underlying structure from data (YouTube videos) and may be learning a lower-dimensional manifold that captures how materials behave. - The conversation pivots to Demis Hassabis’s Nobel Prize lecture conjecture that any pattern generated or found in nature can be efficiently discovered and modeled by a classical learning algorithm. They explore what kinds of patterns or systems might be included: biology, chemistry, physics, cosmology, neuroscience, etc. - AlphaGo and AlphaFold are used as examples of building models of combinatorially high-dimensional spaces to guide search in a tractable way. Hassabis argues that nature’s evolved structures imply learnable patterns, because natural systems have structure shaped by evolutionary processes. This leads to the idea of a potential complexity class for learnable natural systems (LNS) and the possibility that p = NP questions may be reframed as physics questions about information processing in the universe. - They discuss the view that the universe is an informational system, and how that reframes the P vs NP question as a fundamental question about modellability. Hassabis speculates that many natural systems are learnable because they have evolved structure, whereas some abstract problems (like factorizing arbitrary large numbers in a uniform space) may not exhibit exploitable patterns, possibly requiring quantum approaches or brute-force computation. - The dialogue examines whether there could be a broad class of problems that can be solved by polynomial-time classical methods when modeled with the right dynamics and environment—precisely the way AlphaGo and AlphaFold operate. Hassabis emphasizes that classical systems (Turing machines) have already surpassed many expectations by modeling complex biological structures and solving highly challenging tasks, and he believes there is likely more to discover. - They address nonlinear dynamical systems and whether emergent phenomena, such as cellular automata, chaos, or turbulence, might be amenable to efficient classical modeling. Hassabis notes that forward simulation of many emergent systems could be efficient, but chaotic systems with sensitive dependence on initial conditions may be harder to model. He argues that core physics problems, including realistic rendering of physics-like phenomena (e.g., liquids and light interaction), seem tractable with neural networks, suggesting deep structure to nature that can be captured by learning systems. - The conversation shifts to video and world models: Hassabis highlights VOI, video generation, and the hope that future interactive versions could create truly open-ended, dynamically generated game worlds and simulations where players co-create the experience with the environment, beyond current hard-coded or pre-scripted content. They discuss open-world games and the potential for AI to generate content on-the-fly, enabling personalized, ever-changing narratives and experiences. - They discuss Hassabis’s early love of games and his belief that games are a powerful testbed for AI and AGI. He describes the possibility of interactive VO-based experiences that are open-ended and highly responsive to player choices, with emergent behavior that surpasses current procedural generation. - The conversation touches the idea of an open-world world model for AGI: Hassabis imagines a system that can predict and simulate the mechanics of the world, enabling better scientific inquiry and perhaps even a “virtual cell” or virtual biology framework. They discuss AlphaFold as the static prediction of structure and the next step being dynamics and interactions, including protein–protein, protein–RNA, and protein–DNA interactions, and ultimately a model of a whole cell (e.g., yeast). - On the origin of life and origins science: they discuss whether AI could simulate the birth of life from nonliving matter, suggesting a staged approach with a “virtual cell” as a stepping-stone, then moving toward simulating chemical soups and emergent properties that could resemble life. - They consider the nature of consciousness and whether AI systems can or will ever have true consciousness. Hassabis leans toward the view that consciousness (and qualia) may be substrate-dependent and that a classical computer could model the functional aspects of intelligence; but he acknowledges unresolved questions about subjective experience and the potential differences between carbon-based and silicon-based processing. - They discuss the role of AGI in science: the potential for AI to propose new conjectures and hypotheses, to assist in scientific discovery, and perhaps to discover insights that humans might not reach on their own. They acknowledge that “research taste”—the ability to pick the right questions and design experiments meaningfully—is a hard capability for AI to replicate. - They explore the future of video games with AI: Hassabis describes the possibility of open-world, highly interactive experiences that adapt to players’ actions, creating deeply personalized narratives. He compares the future of AI-driven game design to the potential for AI to accelerate scientific progress by modeling complex systems, then translating insights into practical tools and products. - Hassabis discusses the practicalities of running large AI projects at Google DeepMind and Google, noting the balance of startup-like culture with the scale of a large corporation. He emphasizes relentless progress and shipping, while maintaining safety and responsibility, and maintaining collaboration across labs and competitors. - They address data and scaling: Hassabis emphasizes that synthetic data and simulations can help mitigate data scarcity, while real-world data remains essential to guide learning systems. He explains the dynamic between pre-training, post-training, and inference-time compute, noting the importance of balancing improvements across multiple objectives and avoiding overfitting benchmarks. - They discuss governance, safety, and international collaboration: they emphasize the need for shared standards, safety guardrails, and open science where appropriate, while acknowledging the risk of misuse by bad actors and the difficulty of restricting access to powerful AI systems without hampering beneficial applications. Hassabis suggests international cooperation and a CERN-like collaborative model for responsible progress. - They touch on the societal impact of AI: the potential for energy breakthroughs, climate modeling, materials discovery, and fusion, plus the broader economic and political implications. Hassabis anticipates a future where abundant energy reduces scarcity, enabling new levels of human flourishing, but acknowledges distributional concerns and governance challenges. - The dialogue ends with reflections on personal legacies and the human dimension: Hassabis discusses responding to criticism online, his MIT and Drexel affiliations, and the balance between research, podcasting, and public engagement. He emphasizes humility, continuous learning, and openness to collaboration across labs and cultures. Key themes and conclusions preserved from the discussion: - The possibility that many natural patterns are efficiently learnable by classical learning systems if the underlying structure is learned, a view supported by AlphaGo/AlphaFold successes and by phenomena like VOI’s handling of liquids and lighting. - A conjectured link between learnable natural systems and a formal complexity class like LNS, with the broader view that p versus NP is connected to physics and information in the universe. - The potential for classical AI to model complex, nonlinear dynamical systems, including fluid dynamics, with surprising accuracy, given sufficient structure and data. - The idea that nature’s evolutionary processes create patterns that can be reverse-engineered, enabling efficient search and modeling of natural systems. - The role of AI in science as a tool for conjecture generation, hypothesis testing, and accelerating discovery, possibly guiding experiments, reducing wet-lab time, and enabling “virtual cells” and larger-scale simulations. - The interplay between open-world game design, AI-based content creation, and future interactive experiences that adapt to individual players, including the vision of AI-driven world models for AGI. - The practical realities of building and shipping AI products at scale, balancing research breakthroughs with productization, and managing a large organization’s culture and governance to foster safety and innovation. - The ethical and societal questions around AGI: how to ensure safety, how to manage risk from bad actors, the need for international collaboration, governance, and a broad discussion about the role of technology in society. - A hopeful perspective on the long-term future: abundant energy, space exploration, and a transformed civilization driven by AI, with a focus on human values, curiosity, adaptability, and compassion as guiding forces. This summary preserves the essential claims and conclusions of the conversation, including the main positions about learnability, the role of evolution and structure in nature, the potential of classical systems to model complex phenomena, and the broad, multi-domain implications for science, gaming, energy, governance, and society.