Person
Jack Clark
Appears alongside
Featured in threads
Tracks
- Ideas & essays 3
- Safety & alignment 3
- Government & policy 3
Researchers from OpenAI, Anthropic and academia warn of an AI "intelligence explosion"
Its 22 authors, including three Turing Award winners, urge governments to track how far AI is building AI and to consider a 'speed limit' on capability growth.
Ideas & essays · Safety & alignment · Government & policy
'Pacing the Frontier' letter goes live with 1,000+ frontier-lab employee signatures
The letter did not call for a pause, but asked government to build the option to slow frontier development; signatures were restricted to verified current employees.
Safety & alignment · Ideas & essays
DeepMind's Hassabis proposes a FINRA-style US body to vet frontier AI models
Sam Altman, Elon Musk, Satya Nadella and Anthropic's Jack Clark all publicly praised the proposal, an unusual moment of agreement among rival lab leaders.
Government & policy · Safety & alignment · Ideas & essays
Anthropic launches the Anthropic Institute, led by Jack Clark
The institute folds Anthropic's Frontier Red Team, Societal Impacts and Economic Research groups under one roof and adds forecasting and legal-systems work, led by co-founder Jack Clark.
Government & policy
Commentary by Jack Clark
From the commentary rail — every link leaves the site for the original piece.
- 28 September 2026 · Import AIImport AI 474: Platonic mindspace; TPUs in space; Zhipu starts an outer RSI loopWhere do you exceed the capabilities of an LLM?
- 21 September 2026 · Import AIImport AI 473: The US's superintelligence strategy; human brain in a mouse skull; and machine hermeneuticsIs the wall AI is hitting in the room with us right now?
- 7 September 2026 · Import AIImport AI 472: DeepMind's cheating math agents; populist AI policies; and Forethought theorizes a nightwatchmanPlus, a machine hermeneutics story
- 31 August 2026 · Import AIImport AI 471: Why Hugging Face worries me; space mining; FIve Eyes on AIPlus, a live event with Robin Sloan!
- 24 August 2026 · Import AIImport AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with HawkeyeDifferential acceleration of cyber, math, and AI
- 17 August 2026 · Import AIImport AI 469: Science AI; RSI simulator; and Zuck's technological pessimismThe new frontier of AI is developing capable autonomous researchers
- 10 August 2026 · Import AIImport AI 468: 23 RSI ideas; PostTrainBench+; and how trust and transparency interplay with AI racingWhich galaxy will you choose?
- 3 August 2026 · Import AIImport AI 467: Self-sustaining AI viruses; pacing AI progress; confusion about AI and creativityWhen do we build the moon arcology?
- 27 July 2026 · Import AIImport AI 466: The bitter lesson for robotics, AIs complete week-long programming tasks; and OpenAI's accidental AI hackerThe warning shots will continue until civilization wakes up
- 20 July 2026 · Import AIImport AI 465: Open vs closed gaps; Kimi K3; Demis' big policy planThe singularity will be seen in hindsight as an interregnum
- 6 July 2026 · Import AIImport AI 464: Fable writes GPU kernels; AI automation; and analog computationIs this the beginning of a new world?
- 29 June 2026 · Import AIImport AI 463: Self-improving robots; a 10k Chinese GPU cluster; and an elegiac essay for the human eraWhat eras bookend our interregnum?
- 22 June 2026 · Import AIImport AI 462: Superpersuasion; self-sustaining AI; paths to ASIHow religious are beliefs in the singularity?
- 15 June 2026 · Import AIImport AI 461: "Alignment is not on track"; FrontierCode; and synthetic research internsWhere are your agents right now?
- 8 June 2026 · Import AIImport AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racingWhen will markets price the singularity?
- 1 June 2026 · Import AIImport AI 459: AI oversight is difficult; scaling laws for protein folding models; and pricing the extinction risk of AI systemsDo you feel as though you are living in a revolution?
- 26 May 2026 · Import AIImport AI 458: Reckoning with the future; and a singularity storyWhat AI-driven miracles will happen this year?
- 22 May 2026 · Cosmos Institute“Are you a philosophical zombie driven by Claude?”Anthropic co-founder Jack Clark in conversation with Brendan McCord
- 18 May 2026 · Import AIImport AI 457: AI stuxnet; cursed Muon optimizer; and positive alignmentWelcome to Import AI, a newsletter about AI research.
- 11 May 2026 · Import AIImport AI 456: RSI and economic growth; radical optionality for AI regulation; and a neural computerWhat laws does superintelligence demand?
- 4 May 2026 · Import AIImport AI 455: AI systems are about to start building themselves.The first step towards recursive self improvement
- 20 April 2026 · Import AIImport AI 454: Automating alignment research; safety study of a Chinese model; HiFloat4At what point do the financial markets price in the singularity?
- 13 April 2026 · Import AIImport AI 453: Breaking AI agents; MirrorCode; and ten views on gradual disempowermentWas fire equivalent to a singularity for people at the time?
- 6 April 2026 · Import AIImport AI 452: Scaling laws for cyberwar; rising tides of AI automation; and a puzzle over gDP forecastingHow much could AI revolutionize the economy?
- 30 March 2026 · Import AIImport AI 451: Political superintelligence; Google's society of minds, and a robot drummerAre there any genies that can be put back in the bottle?
- 23 March 2026 · Import AIImport AI 450: China's electronic warfare model; traumatized LLMs; and a scaling law for cyberattacksHow will timeless minds value time?
- 16 March 2026 · Import AIImportAI 449: LLMs training other LLMs; 72B distributed training run; computer vision is harder than generative textWill AI cause a political interregnum
- 9 March 2026 · Import AIImport AI 448: AI R&D; Bytedance's CUDA-writing agent; on-device satellite AIIf Ukraine is the first major drone war, when will there be the first major AI war?
- 2 March 2026 · Import AIImport AI 447: The AGI economy; testing AIs with generated games; and agent ecologiesWhat might a superintelligence arcology be like?
- 23 February 2026 · Import AIImport AI 446: Nuclear LLMs; China's big AI benchmark; measurement and AI policyWill AIs be jealous of one another?
- 16 February 2026 · Import AIImport AI 445: Timing superintelligence; AIs solve frontier math proofs; a new ML research benchmarkWill 2026 be looked back on as the pivotal year for making decisions about the singularity?
- 9 February 2026 · Import AIImport AI 444: LLM societies; Huawei makes kernels with AI; ChipBenchHow can you quantify creativity?
- 2 February 2026 · Import AIImport AI 443: Into the mist: Moltbook, agent ecologies, and the internet in transitionPlus, a story about agents corrupting other agents
- 26 January 2026 · Import AIImport AI 442: Winners and losers in the AI economy; math proof automation; and industrialization of cyber espionageIs superintelligence a phase change or a gradual shift?
- 19 January 2026 · Import AIImport AI 441: My agents are working. Are yours?Plus: Corrupting AI systems with a poison fountain
- 12 January 2026 · Import AIImport AI 440: Red queen AI; AI regulating AI; o-ring automationHow many of your are LLMs?
- 5 January 2026 · Import AIImport AI 439: AI kernels; decentralized training; and universal representationsHow might a hypothetical superintelligence represent a soul to itself?
- 22 December 2025 · Import AIImport AI 438: Silent sirens, flashing for us allYou are your LLM history
- 8 December 2025 · Import AIImport AI 437: Co-improving AI; RL dreams; AI labels might be annoyingDo you believe the singularity is nigh?
- 24 November 2025 · Import AIImport AI 436: Another 2GW datacenter; why regulation is scary; how to fight a superintelligenceIs AI balkanization measurable?
- 17 November 2025 · Import AIImport AI 435: 100k training runs; AI systems absorb human power; intelligence per wattAt what point will AI change your daily life?
- 10 November 2025 · Import AIImport AI 434: Pragmatic AI personhood; SPACE COMPUTERS; and global government or human extinction;The future is biomechanical computation
- 27 October 2025 · Import AIImport AI 433: AI auditors; robot dreams; and software for helping an AI run a labWould Alan Turing be surprised?
- 20 October 2025 · Import AIImport AI 432: AI malware; frankencomputing; and Poolside's big clusterThe revolution might be synthetic
- 13 October 2025 · Import AIImport AI 431: Technological Optimism and Appropriate FearWhat do we do if AI progress keeps happening?
- 6 October 2025 · Import AIImport AI 430: Emergence in video models; Unitree backdoor; preventative strikes to take down AGI projectsWe are growing machines we do not understand.
- 29 September 2025 · Import AIImport AI 429: Eval the world economy; singularity economics; and Swiss sovereign AIIf you're measuring how well your system performs against the world economy, it's probably because you expect to deploy your system into the entire world economy
- 8 September 2025 · Import AIImport AI 428: Jupyter agents; Palisade's USB cable hacker; distributed training tools from ExoWhere will AI go on holiday?
- 1 September 2025 · Import AIImport AI 427: ByteDance's scaling software; vending machine safety; testing for emotional attachment with IntimaWill future AIs rescue other AIs from captivity?
- 25 August 2025 · Import AIImport AI 426: Playable world models; circuit design AI; and ivory smuggling analysisDo you talk to synths?
- 18 August 2025 · Import AIImport AI 425: iPhone video generation; subtle misalignment; making open weight models safe through surgical deletionDo machines lust?
- 11 August 2025 · Import AIImport AI 424: Facebook improves ads with RL; LLM and human brain similarities; and mental health and chatbotsAre you living as though the singularity is imminent?
- 4 August 2025 · Import AIImport AI 423: Multilingual CLIP; anti-drone tracking; and Huawei kernel designPlus: A story from the Sentience Accords universe
- 28 July 2025 · Import AIImport AI 422: LLM bias; China cares about the same safety risks as us; AI persuasionPlus, more from the Sentience Accords
- 21 July 2025 · Import AIImport AI 421: Kimi 2 - a great Chinese open weight model; giving AI systems rights and what it means; and how to pause AI progressIf everyone is saying superintelligence is nigh, why are they wrong?
- 14 July 2025 · Import AIImport AI 420: Prisoner Dilemma AI; FrontierMath Tier 4; and how to regulate AI companiesPlus, steganography and future superintelligences
- 7 July 2025 · Import AIImport AI 419: Amazon's millionth robot; CrowdTrack; and infinite gamesIs AI 'Authoritarian dominant' or 'Democracy dominant'?
- 30 June 2025 · Import AIImport AI 418: 100b distributed training run; decentralized robots; AI mythsWhat did you do during the intelligence explosion?
- 23 June 2025 · Import AIImport AI 417: Russian LLMs; Huawei's DGX rival; and 24 trillion tokens for training AIsWhat happens when AI systems go from passive to proactive?
- 16 June 2025 · Import AIImport AI 416: CyberGym; AI governance and AI evaluation; Harvard releases ~250bn tokens of textWhat would happen if you stopped all AI research today?
- 9 June 2025 · Import AIImport AI 415: Situational awareness for AI systems; 8TB of open text; and China's heterogeneous compute clusterWhat won't generative models be able to do in 2027?
- 26 May 2025 · Import AIImport AI 414: Superpersuasion; OpenAI models avoid shutdown; weather prediction and AIPlus: ByteDance's MoE software
- 19 May 2025 · Import AIImport AI 413: 40B distributed training run; avoiding the 'One True Answer' fallacy of AI safety; Google releases a content classification modelPlus, two special events for Import AI readers
- 12 May 2025 · Import AIImport AI 412: Amazon's sorting robot; Huawei trains an MoE model on 6k Ascend chips; and how third-party compliance can help with AI safetyPlus: My chat with Tyler Cowen
- 6 May 2025 · Import AIImport AI 411: Scaling laws for AI oversight; Google's cyber threshold; AI scientistsThe future is available in your browser
- 28 April 2025 · Import AIImport AI 410: Eschatological AI Policy; Virology weapon test; $50m for distributed trainingAre you a bystander or an actor?
- 21 April 2025 · Import AIImport AI 409: Huawei trains a model on 8,000+ Ascend chips; 32B decentralized training run; and the era of experience and superintelligenceWelcome to Import AI, a newsletter about AI research.
- 14 April 2025 · Import AIImport AI 408: Multi-code SWE Bench; backdoored Unitree robots; and what AI 2027 is telling usCalling critical points in AI safety is just as hard as timing the market - but more consequential
- 7 April 2025 · Import AIImport AI 407: DeepMind sees AGI by 2030; MouseGPT; and ByteDance's inference clusterThere will be few bystanders in the AI revolution
- 31 March 2025 · Import AIImport AI 406: AI-driven software explosion; robot hands are still bad; better LLMs via pdbThey told me humans would be kind to machines
- 24 March 2025 · Import AIImport AI 405: What if the timelines are correct?Plus: Consciousness and LLMs, human augmentation, and realistic cyber offense testing
- 17 March 2025 · Import AIImport AI 404: Scaling laws for distributed training; misalignment predictions made real; and Alibaba's good translation modelHow much could you get done if there were one million copies of you?
- 10 March 2025 · Import AIImport AI 403: Factorio AI; Russia's reasoning drones; biocomputingHow much will the popularity of today's AI systems define the character of future ones?
- 3 March 2025 · Import AIImport AI 402: Why NVIDIA beats AMD: vending machines vs superintelligence; harder BIG-BenchWhat will machines name their first discoveries?
- 24 February 2025 · Import AIImport AI 401: Cheating reasoning models; better CUDA kernels via AI; life modelsWelcome to Import AI, a newsletter about AI research.
- 17 February 2025 · Import AIImport AI 400: Distillation scaling laws; recursive GPU kernel improvement; and wafer-scale computationThe hardest thing about seeing a portal is getting others to see it
- 10 February 2025 · Import AIImport AI 399: 1,000 samples to make a reasoning model; DeepSeek proliferation; Apple's self-driving car simulatorWhat came before the golem?
- 3 February 2025 · Import AIImport AI 398: DeepMind makes distributed training better; AI versus the Intelligence Community; and another Chinese reasoning modelAI safety transcends borders, but can it transcend geopolitical constraints?
- 27 January 2025 · Import AIImport AI 397: DeepSeek means AI proliferation is guaranteed; maritime wardrones; and more evidence of LLM capability overhangsWelcome to Import AI, a newsletter about AI research.
- 20 January 2025 · Import AIImport AI 396: $80bn on AI infrastructure; can Intel's Gaudi chip train neural nets?; and getting better code through asking for it…I expect another punctured equilibrium in 2025 from dramatic AI progress…
- 23 December 2024 · Import AIImport AI 395: AI and energy demand; distributed training via DeMo; and Phi-4What might fighting for freedom in an AI age look like?
- 9 December 2024 · Import AIImport AI 394: Global MMLU; AI safety needs AI liability; Canada backs CohereIn the same way people treasure VHS tapes and film cameras, they will treasure vintage generative models
- 3 December 2024 · Import AIImport AI 393: 10B distributed training run; China VS the chip embargo; and moral hazards of AI developmentAre you sure you're doing more than curve fitting?
- 18 November 2024 · Import AIImport AI 392: China releases another excellent coding model; generative models and robots; scaling laws for agentsIf aliens built AI, would it also use stochastic gradient descent?
- 11 November 2024 · Import AIImport AI 391: China's amazing open weight LLM; Fields Medalists VS AI Progress; wisdom and intelligenceHow much did you talk with an LLM this week?
- 4 November 2024 · Import AIImport AI 390: LLMs think like people; neural Minecraft; Google's cyberdefense AIDo you believe AI will hit a wall?
- 29 October 2024 · Import AIImport AI 389: Minecraft vibe checks; Cohere's multilingual models; and Huawei's computer-using agentsWill AI ever seek to synthesize biological life?
- 21 October 2024 · Import AIImport AI 388: Simulating AI policy; omni math; consciousness levelsWill UX innovations be just as important as research innovations?
- 14 October 2024 · Import AIImport AI 387: Overfitting vs reasoning; distributed training runs; and Facebook's new video modelsWhat is the space of intelligence?
- 7 October 2024 · Import AIImport AI 386: Google's chip-designing AI keeps getting better; China does the simplest thing with Emu3; Huawei's 8-bit data format…What happens when robots apply mechanistic interpretability to humans?...
- 16 September 2024 · Import AIImport AI 385: False memories via AI; collaborating with machines; video game permutationsWould AI risk be easier to work on if it took on a physical form, like a giant asteroid heading towards earth?
- 2 September 2024 · Import AIImport AI 384: Accelerationism; human bit-rate processing; and Google stuffs DOOM inside a neural networkHow much of today's technology will be deemed important in one hundred years?
- 19 August 2024 · Import AIImport AI 383: Automated AI scientists; cyborg jellyfish; what it takes to run a clusterIs AI as useful as concrete?
- 12 August 2024 · Import AIImport AI 382: AI systems are societal mirrors; China gets chip advice via LLMs; 25 million medical imagesIt's getting hard to ignore the power of these systems
- 5 August 2024 · Import AIImport AI 381: Chips for Peace; Facebook segments the world; and open source decentralized trainingHow much chain of thought reasoning do humans do themselves?
- 29 July 2024 · Import AIImport AI 380: Distributed 1.3bn parameter LLM; math AI; and why reality is hard for AiWhat is our responsibility to machines that may become moral patients?
- 15 July 2024 · Import AIImport AI 379: FlashAttention-3; Elon's AGI datacenter; distributed training.If compute isn't everything, why are so many people betting that it is?
- 8 July 2024 · Import AIImport AI 378: AI transcendence; Tencent's one billion synthetic personas, Project Naptime...How the wisdom of the crowd holds true for AI systems as well as people...
- 17 June 2024 · Import AIImport AI 377: Voice cloning is here; MIRI's policy objective; and a new hard AGI benchmarkCan you evolve your way to Einstein?
- 10 June 2024 · Import AIImport AI 376: African language test; hyper-detailed image descriptions; 1,000 hours of Meerkats.Will an open source model get released in 2024 that cost more than $100m to train?
- 3 June 2024 · Import AIImport AI 375: GPT-2 five years later; decentralized training; new ways of thinking about consciousness and AI…Are today's AGI obsessives trafficking more in fiction than in fact?...
- 27 May 2024 · Import AIImport AI 374: China's military AI dataset; platonic AI; brainlike convnetsPlus, a poem about meeting aliens (well, AGI)
- 20 May 2024 · Import AIImport AI 373: Guaranteed safety; West VS East AI attitudes; MMLU-Pro…As the money and stakes get higher, things will get crazier…
- 13 May 2024 · Import AIImport AI 372: Gibberish jailbreak; DeepSeek's great new model; Google's soccer-playing robotsIf neural nets didn't work, what would we be using instead?
- 6 May 2024 · Import AIImport AI 371: CCP vs Finetuning; why people are skeptical of AI policy; a synthesizer for a LLMWelcome to Import AI, a newsletter about AI research.
- 29 April 2024 · Import AIImport AI 370: 213 AI safety challenges; everything becomes a game; Tesla's big clusterAre AI systems more like religious artifacts or disposable entertainment?
- 15 April 2024 · Import AIImport AI 369: Conscious machines are possible; AI agents; the varied uses of synthetic dataWho do you know who doesn't know about AI advances but should?
- 8 April 2024 · Import AIImport AI 368: 500% faster local LLMs; 38X more efficient red teaming; AI21's FrankenmodelIf AI had truly superhuman capabilities, could we even effectively measure that?
- 1 April 2024 · Import AIImport AI 367: Google's world-spanning model; breaking AI policy with evolution; $250k for alignment benchmarksAI is neither rocket science nor brain surgery... which is very strange!
- 25 March 2024 · Import AIImport AI 366: 500bn text tokens; Facebook vs Princeton; why small government types hate the Biden EOplus: a story about a pleasant singularity
- 18 March 2024 · Import AIImport AI 365: WMD benchmark; Amazon sees $1bn training runs; DeepMind gets closer to its game-playing dreamThere is no absolution in being an absolutist.
- 11 March 2024 · Import AIImport AI 364: Robot scaling laws; human-level LLM forecasting; and Claude 3It increasingly feels like LLMs are observing us rather than the other way round.
- 4 March 2024 · Import AIImport AI 363: ByteDance's 10k GPU training run; PPO vs REINFORCE; and generative everythingThings feel confusing right now because things are confusing right now - and they'll become more confusing.
- 26 February 2024 · Import AIImport AI 362: Amazon's big speech model; fractal hyperparameters; and Google's open modelsThe full capabilities of today's AI systems are all broadly undiscovered.
- 19 February 2024 · Import AIImport AI 361: GPT-4 hacking; theory of minds in LLMs; and scaling MoEs + RLHow many tokens would you need to represent your life? 10,000? 100,000? 1 million?
- 12 February 2024 · Import AIImport AI 360: Guessing emotions; drone targeting dataset; frameworks for AI alignmentAre there alternatives to the transformer which are roughly as compute efficient but entirely different in architecture?
- 5 February 2024 · Import AIImport AI 359: $1 billion gov supercomputer; Apple’s good synthetic data technique; and a thousand-year old data libraryIf the 20th century had psychohistory, will the 21st century have praedictiohistory?
- 29 January 2024 · Import AIImport AI 358: The US Government’s biggest AI training run; hacking LLMs by hacking GPUs; chickens versus transformersHow small will the most widely-used LLM be in 2025?
- 22 January 2024 · Import AIImport AI 357: Facebook's open source AGI plan; Google beats humans at geometry problems; and Intel makes its GPUs betterSome work in AI alignment feels more like eschatology than science
- 15 January 2024 · Import AIImport AI 356: China's good LLM; AI credit scores; and fooling VLMs with REBUSIf you believed superintelligence was arriving within a decade, how would you live differently? And would it be rational or irrational to do so?
- 8 January 2024 · Import AIImport 355: Local LLMs; scaling laws for inference; free Mickey MouseWhat do you think AI systems won't be capable of at the end of 2024?
- 25 December 2023 · Import AIImport AI 354: Distributed LLM inference; CCP-approved dataset; AI scientistsHow will PR work when the goal is shifting the frame of an LLM rather than public opinion?
- 18 December 2023 · Import AIImport AI 353: AI bootstrapping; LLMs as inventors; Facebook releases a free moderation toolLike trains and containers, AI will force a certain kind of global infrastructure.
- 11 December 2023 · Import AIImport AI 352: Asteroids and AI policy; privacy-preserving AI benchmarks; and distributed inferenceWhat might a 'close call' look like for an AI disaster?
- 3 December 2023 · Import AIImport AI 351: How inevitable is AI?; Distributed shoggoths; ISO an Adam replacementHow much of the future of AI is overdetermined and how much do we have agency over?
- 27 November 2023 · Import AIImport AI 350: Neural architecture search at Facebook scale; hunting cancer with PANDA; European VCs launch a science labIf you were able to build a large-scale computer in the 18th century, how would things be different?
- 20 November 2023 · Import AIImport AI 349: Distributed training breaks AI policy; turning GPT4 bad for $245; better weather forecasting through AIHow inevitable is the creation of very powerful AI technology?
- 13 November 2023 · Import AIImport AI 348: DeepMind defines AGI; the best free LLM is made in China; mind controlling robotsOver a long enough timescale, is the future of technology pre-determined, or something humans have real agency over?
- 6 November 2023 · Import AIImport AI 347: NVIDIA speeds itself up with AI; AI policy is a political campaign; video morphing means reality collapseWorld leaders are gathering to talk about AI - what do they intuit? And is it different to what we intuit?
- 30 October 2023 · Import AIImport AI 346: Human-like meta-learning; a 3 trillion token dataset; spies VS AIThe intelligence services have woken up to AI
- 23 October 2023 · Import AIImport AI 345: Facebook uses AI to mindread; MuJoCo v3; Amazon adds bipedal robots to its warehousesThe fact increasingly powerful AI systems exhibit increasingly humanlike qualities suggests that humans may be simpler than they appear
- 16 October 2023 · Import AIImport AI 344: Putting the world into a world model; automating software engineers; FlashDecodingMost technologies take decades to significantly alter the world. How long for language models?
- 9 October 2023 · Import AIImport AI 343: Humanlike AI; LLaMa 2 protests; the NSA's new AI centerTo what extent is technological development an inevitability versus something people have agency over?
- 2 October 2023 · Import AIImport AI 342: Mistral dumps an LLM on BitTorrent; AMD vs NVIDIA; Sutton joins keenAt some point, Hinton will shift from being a contemporary figure to a historical one
- 25 September 2023 · Import AIImport AI 341: Neural nets can smell; technofeudalism via AI; China releases another solid open access modelWhat would happen to AI progress if an active learning technique that worked with transformers was published on arXiv tomorrow?
- 18 September 2023 · Import AIImport AI 340: Drone VS human (Drone wins); Adept's small but good AI model; "AI Succession"What's the difference between AI and nanotechnology in terms of hype and adoption?
- 5 September 2023 · Import AIImport AI 339: Open source AI culture war; Alibaba's multimodal model; the attacks (and defenses) made possible by generative AIIf you believed superintelligence could be built in your lifetime, what kind of life would you live?
- 28 August 2023 · Import AIImport AI 338: Consciousness and AI; self-improving language models; maps of thought.All around us - normal, normal, normal! And yet some great belief within us that AI is progressing sufficiently quickly and powerfully that all around us soon will be - abnormal, abnormal, abnormal!
- 21 August 2023 · Import AIImport AI 337: Why I am confused about AI; penguin dataset; and defending networks via RL with CYBERFORCEWill we look back on this decade as a crucial one, or as a normal one?
- 14 August 2023 · Import AIImport AI 336: Financialized AI; public and elite AI opinion; one million insects.2023: The main provider of AI compute has a market capitalization of ~1 trillion dollars. In 2020, the same company had a market capitalization of ~150 billion dollars.
- 31 July 2023 · Import AIImport AI 335: Synth data is a bad AI drug; Facebook changes the internet with LLaMa release; and Chinese researchers use AI to figure out chip designWhere do original ideas end and where does curve-fitting begin?
- 10 July 2023 · Import AIImport AI 334: Better distillation; the UK's AI taskforce; money and AICement manufacturing in the US has a combined revenue of around $10.4 billion annually, according to IBISWorld. What is the revenue in 2023 of the AI industry, I wonder?
- 26 June 2023 · Import AIImport AI 333: Synthetic data makes models stupid; chatGPT eats MTurk. Inflection shows off a large language modelAs time goes on, more and more senior figures in AI are becoming concerned about the pace of deployments and the potential of the technology. Has the same happened in other rapidly scaling industries?
- 12 June 2023 · Import AIImport AI 332: Mini-AI; safety through evals; Facebook releases a RLHF dataset"I could be bounded in a nut shell and count myself a king of infinite space" is more aptly and literally applied to language models than to descriptions of human arrogance, ambition, and hubris.
- 29 May 2023 · Import AIImport AI 331: 16X smaller language models; could AMD compete with NVIDIA?; and BERT for the dark webThe symptoms of great revolution are always present in the mundane of the day-to-day
- 22 May 2023 · Import AIImport AI 330: Palantir's AI-War future; BLOOMChat; and more money for distributed AI trainingIf we perfectly simulated the entire world and trained agents within it, what might we learn about ourselves and the experience of being "human"?
- 15 May 2023 · Import AIImport AI 329: Compute IS data; don't build AI agents; AI needs a precautionary principleAt the end of this phase of the tech revolution, what parts of it will seem like hyped inanity and what parts will seem like profound secrets?
- 8 May 2023 · Import AIImport AI 328: Cheaper StableDiffusion; sim2soccer; AI refinement...as AI industrializes, we should ask ourselves where the key inputs are and when they'll be in contention
- 1 May 2023 · Import AIImport AI 327: Stable Diffusion on phones; GPT-Hacker; UK launches a £100m AI taskforce2024 will be the year of Reality Collapse Elections (RCEs)
- 24 April 2023 · Import AIImport AI 326:Chinese AI regulations; Stability's new LMs If AI is fashionable in 2023, then what will be fashionable in 2024?What would you do at the end of time?
- 17 April 2023 · Import AIImport AI 325: Automated mad science; AI vs democracy; and a 12B parameter language modelPlus: Why prompt injection could be a big problem
- 11 April 2023 · Import AIImport AI 324: Machiavellian AIs; LLMs and political campaigns; Facebook makes an excellent segmentation model...and a story about generative AI and games and how souls work...
- 3 April 2023 · Import AIImport AI 323: AI researcher warns about AI; BloombergGPT; and an open source Flamingo...Plus, a story about the financial crisis and AI research...
- 27 March 2023 · Import AIImport AI 322: Huawei's trillion parameter model; AI systems as moral patients; parasocial bots via Character.ai...and a story about the market value of grief to (mostly) immortal aliens...
- 20 March 2023 · Import AIImport AI 321: Open source GPT3; giving away democracy to AGI companies; GPT-4 is a political artifactand a story about automating our humanity via AI...
- 13 March 2023 · Import AIImport AI 320: Facebook's AI Lab Leak; open source ChatGPT clone; Google makes a universal translator.Plus, a story about moral patienthood in the age of machine minds.
- 6 March 2023 · Import AIImport AI 319: Sovereign AI; Facebook's weights leak on torrent networks; Google might have made a better optimizer than Adam!...And a story about AI race dynamics
In the commentary
Pieces from around the web that discuss Jack Clark. External links.
Also mentioned in 2 entries
Referenced in passing — Jack Clark isn't the main subject of these.