Loading activity…
@chrisalbon
The research and practice of generating knowledge with the help of AI at http://machinesofknowledge.com | Senior Director, ML & Data @Wikimedia
AI
Anthropic reports 17 of 211 lab-tested designs met binding criteria, but none were tested for toxin neutralization.
AI
A poster argues that judgment and taste will remain hard for AI to master across domains and over time.
AI
A user says the chart comes to mind whenever someone on the app, including themselves, predicts AI's future.
AI
A user suggests 10× the input buys roughly a constant increment of output in both relationships.
AI
A user argues the term now covers software written mostly by an agent, not just quick, casual projects.
AI
A user recalls someone spending a month alone reading ML research and says they still want to try it.
AI
A user quotes a cliff-edge analogy for radiologists and says they're glad people didn't listen.
AI
A reader says he expected to skim Ramez Naam’s essay but read it twice and took notes.
AI
A user offers the line as a one-sentence challenge in a quote post.
AI
A user jokes that replacing the toilet paper roll is 'infinite alpha.'
AI
A user bluntly demands its return, ending the appeal with “you cowards.”
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 people with high p(doom) can imagine AI with whatever capabilities would make their fears come true.
AI
Kevin Roose says the book became a New York Times bestseller in its first week.
AI
A user endorses local AI for fun but rejects an appeal to 'liberate' users from OpenAI and Anthropic.
AI
A user bets it could improve AI alignment but worries it might make a model's persona paranoid and excessively scrupulous.
AI
OpenAI says its models helped identify the employee who planted the device; the FBI made an arrest.
AI
A user argues AI will help engineers add more value, but expects private-equity-owned firms to squeeze every dollar.
AI
A user says something Lewis wrote 78 years earlier captures exactly how they feel about AI.
AI
A user suspects a link to Nous Research's Hermes Agent but says they can't confirm it.
