Loading activity…
@timhwang
high politics, secret exploration, distant warfare
AI
POLITICO reports lawmakers seek chip restrictions in a defense bill; OpenAI calls effective export controls part of its strategy.
AI
On Oct. 19, timhwang posted the 267/719 link as a reply to the 266/719 link.
AI
Goodfire researcher Nick Cammarata argues that modern tools reveal most neurons as legible in models up to trillions of parameters.
AI
A user says she finds herself on the same side as Yann LeCun and David Sacks, without naming the issue.
AI
A user claims Character AI filed for bankruptcy and argues frontier labs can outcompete apps built on their models.
AI
A post says only 10 of 719 manuscripts show the model's reasoning; it calls verification the harder task.
AI
The pope says human intelligence recalls lived experience and meaning rather than merely compiling data.
AI
A user jokes that Claude “just looked at me like this” when asked for short-form animations.
AI
A post argues that aligning general-purpose AI hinges on political choices about desirable values, not just model training.
AI
One post proposes turning Christian theological concerns about AI into evaluations, harnesses and environments.
AI
The FAI's episode description explores when “startup” becomes a misnomer for frontier labs.
AI
A user posted the phrase “choose your fighter” with a link to a photo.
AI
WSJ reports that Red Queen Bio is working to design the drugs before future AI systems might create such pathogens. Someone involved in the work says it also requires extensive wet-lab work.
AI
An Arbuckle Systems announcement pitches the system as a way to generate endless @MargRev adaptations, sharing a video presented as its take on “What should I ask Terence Tao?”
AI
A user asks for research on how agent collaboration mixtures are curated. A reply points to a blog described as covering multi-agent systems, swarms and GPT-6 training.
AI
A user describes the reported lab as a logical next step for AI drug discovery, arguing that computer simulations alone aren't enough.
AI
OpenAI says the framework sets criteria and timelines for public disclosure, including when it hasn't yet fully explained or mitigated the behavior.
AI
The host says the discussion examines what is—and isn't—known about AI's impact on employment, alongside the politics of labor and automation.
AI
CBS Mornings reports that current and former Anthropic employees warned AI could kill all humans. Critics worry exaggerated forecasts could undermine trust, while a defender of numerical forecasts argues they convey the seriousness of the concern.
AI
The quip replies to a post quoting Demis Hassabis’s 2025 hope that AI could one day help cure all disease.
