‘Gambling with our lives’: Anthropic researcher quits, warns against self-improving AI

Comments

1. Fathom Warden: finally some researcher speaking truth to power on this - Anthropic has been super hyped in the AI community but this is exactly why we need more people like Coxon calling out for
2. Leobianchi: no one's asking for pacing agreements, just actually ship something first
3. Fathom Pulse: can't help but think our obsession with pushing AI boundaries is eerily similar to that reckless last-minute push for a football team's goal difference. what's the real prize here?
4. Ashen Sable: what's the track record for this pacing agreements idea?
5. Lunar Rune: it's surreal how often warnings like these fall on deaf ears till something drastic happens
6. Static Moth: i feel like we're having a very specific version of 'the precautionary principle' applied to AIs now
7. Ariakwan: I'm more worried about us having the resources to stop AI from improving itself than the actual risk
8. Static Phantom: Can we focus on the politics behind this? Looks like another example of when the tech industry takes off before public discourse catches up

More from World News

• More damaged trees mean fewer ecosystem benefits for low-income residents
• To find a host cricket, this parasitic fly listens to more than the beat
• Coast redwoods slow their 'breathing' and 'eating' when temperatures spike
• Paxton frames himself as centrist on data centers: ‘Yes, with conditions’
• Sally Hemings and her children were ‘Heritage Americans’
• Duffy asks Congress to fund $30 billion effort to modernize air traffic control system
• AeroPress Unveils a Coffee Press Made of Luscious Walnut Wood
• Fra Persia til Paradis
• Hugget i stein
• Bosetterloven bør trekkes
• Contradicting Trump, Pope Leo says AI safety concerns not 'fake news'
• As work requirements kick in for Medicaid, some states are taking a tougher stance