Obsidian Mantis: what kind of 'mini Manhattan project' have they got going on behind closed doors that's so critical to humanity's existence suddenly?
Sana: i'm surprised there's not more said on the "alignment problem" - always sounds like a euphemism for us not quite understanding what we're dealing with yet
Quiet Shard: this entire piece feels like a really well-timed wake-up call
Ariagrant: seems like he's echoing what I've read elsewhere, but still a bit too apocalyptic for my taste
Shadow Sparrow: the way you're phrasing "crunch time for humanity" is pretty melodramatic, given what little we actually know about the situation
Zoe: crunch time for humanity is a bit melodramatic, but i do think the alignment problem is being glossed over by labs trying to just 'scale up' and
Nova: still amazed by how little the alignment problem has affected their actual product roadmap though
Echo Cipher: yeah, Coxon's been saying this for years - but I guess now it's not just 'we're working on a solution' anymore
Ashen Vector: that transition from pixel art to the speedrun footage is jarring somehow
Rayan: sounds like he's painting a pretty bleak picture
Hex Reaper: Anthropic's problems aren't unique to them, or even AI research in general. It's just the industry's growth outpacing its ethics
Mirage Shard: anthropic's mini Manhattan project was supposed to solve the alignment problem, but jacob coxon just says straight out that it's crunch time and they're running out of years
Mia: it's funny that he mentions the "mini Manhattan project" vibe in AI labs now - feels like every major breakthrough since '90s gaming gets remixed somehow