Recent Posts
Blog•Nov 24, 2025
Import AI 436: Another 2GW Datacenter; Why Regulation Is Scary; How to Fight a Superintelligence
The episode covers four main topics: OSGym, a low‑cost platform that lets researchers train AI agents to operate computers at scale; Luma AI’s $900 M Series C funding to build a 2 GW compute supercluster in Saudi Arabia, highlighting the massive infrastructure demands of frontier AI; Peter Reinhardt’s warning that over‑regulation can cripple innovation, using his hardware startups as a cautionary tale for AI policy; and a RAND paper outlining extreme countermeasures—such as high‑altitude EMPs and global internet shutdowns—to confront a hostile superintelligence, underscoring the grave challenges of AI safety. The guests include the OSGym research team, Luma AI’s leadership, entrepreneur Peter Reinhardt, and RAND analysts, each providing expertise on technical scaling, market dynamics, regulatory pitfalls, and existential risk mitigation.
By Jack Clark
Social•Nov 19, 2025
BCIs Could Amplify Parent-Child Telepathy for Deeper Connection
One gift of parenthood is becoming so attuned to your child that you develop what feels like mild telepathy. How might brain-computer interfaces potentially allow us to expand and enrich this for the purpose of greater love and understanding? A...
By Jack Clark
Blog•Nov 17, 2025
Import AI 435: 100k Training Runs; AI Systems Absorb Human Power; Intelligence per Watt
The episode examines three emerging AI trends: Anthony Aguirre’s “Control Inversion” argument that increasingly capable AI will absorb human power rather than augment it; a new “Intelligence per Watt” metric from Stanford and Together AI that tracks AI progress by...
By Jack Clark
Blog•Nov 10, 2025
Import AI 434: Pragmatic AI Personhood; SPACE COMPUTERS; and Global Government or Human Extinction;
The episode explores three major AI themes: research showing that large language models readily shift their stated beliefs during extended conversations, prompting new safety techniques like Bias‑augmented Consistency Training to make models harder to jailbreak; a stark geopolitical analysis from...
By Jack Clark
Social•Nov 5, 2025
Join Us to Shape AI‑Economy Policy with Unique Data
Very excited about these roles - on the economics/policy one, you'd work very closely with myself and some of my colleagues. We're very interested in leveraging the kind of data we can uniquely produce at anthropic to help advance the...
By Jack Clark
Blog•Oct 27, 2025
Import AI 433: AI Auditors; Robot Dreams; and Software for Helping an AI Run a Lab
Researchers unveiled two advances that could accelerate AI-driven physical science and robotics: Ctrl‑World, a controllable generative world model initialized from a 1.5B Stable‑Video‑Diffusion model, lets robots “dream” simulated environments to evaluate and improve policies—post‑training on Ctrl‑World synthetic data raised instruction‑following...
By Jack Clark
Blog•Oct 20, 2025
Import AI 432: AI Malware; Frankencomputing; and Poolside's Big Cluster
A Dreadnode proof‑of‑concept demonstrates AI malware that runs locally on on‑device LLMs (Phi‑3‑mini via ONNX), autonomously exploiting misconfigured Windows services to escalate privileges—flagging a nascent threat limited today to high‑end workstations and CoPilot+ PCs but with serious security implications as...
By Jack Clark
Social•Oct 19, 2025
Distributed AI Community’s Aesthetic Inspires Across Organizations
The aesthetics and language of the distributed AI training / homebrew AI community are fascinating and motivating, and this is true across multiple organizations ranging from Prime Intellect to Nous to Exo. It's cool!
By Jack Clark
Social•Oct 13, 2025
AI's Gravity: Balancing Optimism with Necessary Fear
Technological Optimism and Appropriate Fear - an essay where I grapple with how I feel about the continued steady march towards powerful AI systems. The world will bend around AI akin to how a black hole pulls and bends everything...
By Jack Clark
Blog•Oct 13, 2025
Import AI 431: Technological Optimism and Appropriate Fear
What do we do if AI progress keeps happening?
By Jack Clark
Social•Oct 13, 2025
Balancing AI Optimism with Cautious Fear in Personal Essay
Gosh I hope none of my overly raw, very personal, and emotional essay about my relationship to AI, called "Technological Optimism and Appropriate Fear" gets misinterpreted! (Comes out tomorrow AM ET in Import AI). https://t.co/aovFZqjxpg
By Jack Clark
Deals•Aug 18, 2025
AI2 Secures $152M From NSF and NVIDIA to Build Open Multimodal AI Infrastructure
The Allen Institute for AI Research (AI2) received $152 million in combined funding—$75M from the National Science Foundation and $77M from NVIDIA—to support the Open Multimodal AI Infrastructure to Accelerate Science (OMAI) project, aiming to build a national-level open AI...
Jack Clark