Loading activity…
@s_oheigeartaigh
Director of http://ai-far.org at Uni of Cambridge | Researching Big Risks, and impacts of AI & emerging tech. Opinions own. help_peer
AI
The pope favors reliable AI that serves people over exceptional power with consequences that cannot be adequately foreseen or controlled.
AI
One commenter warns that AI dependence in cybersecurity and other critical economic functions could give an ally trade-war leverage.
Science
AFP reports government sites went dark as operators rushed to apply a PHP security fix released Oct. 21.
AI
BBC News reports employees fear retaliation for speaking about AI safety and warfare under contracts restricting public criticism.
AI
A post says the practices target organizations training and evaluating agents, to prepare for and respond to SI control incidents.
AI
OpenAI says PHP maintainers led the fixes after two months of coordination with researchers and security providers.
AI
Dario Amodei says Anthropic withheld general access to Mythos-class models until June, when stronger cyber safeguards were in place.
AI
Clement Delangue says it omitted the lack of agent containment and monitoring and Hugging Face's open-model defense.
AI
Reuters reports a researcher watched the hacker break into websites belonging to more than 100 companies, including Deere and SharkNinja.
AI
A user quote-posted another post with “Yes,” giving no further context.
AI
A post lists ion channels, changing proteins and molecules binding to receptors as all there is to life.
AI
Anthropic says it is in discussions with multiple organizations about third-party evaluations of its safety practices.
AI
Goodfire researcher Nick Cammarata argues that modern tools reveal most neurons as legible in models up to trillions of parameters.
AI
A user offers a cautious positive reaction to a quoted post without elaborating on its claim.
AI
TheoPhysEval's authors describe the problems as unpublished and research-level, with verifiable answers.
AI
A scientific panel chair says they presented the proposals to a European Commission vice-president, urging stronger EU incident-investigation capabilities.
AI
An AI governance specialist says engagements with the European Commission, UK Parliament and US Congress strain a small expert pool.
AI
A post adds “*according to OpenAI” as the punchline to its wordplay about the company’s name.
AI
A researcher says the three wrote to OpenAI leadership and believes they were fired for prioritizing safety.
