Joe Carlsmith

There's a list of bullet points here.

Joe Carlsmith, building AIs that do human-like philosophy considers philosophy more broadly rather than just moral philosophy and usefulness for alignment specifically rather than usefulness more broadly.

Matthew Adelstein, aka Bentham's Bulldog with eighty k. Mitigating extinction risk isn't the only reason to work on AI section on moral errors and less closely Adelstein with forethought.

Can AI do philosophy? Wei Dai has written a bunch about related ideas of automating philosophy, though often my perspective is a bit different than his, I think.

But the people behind Anthropic have been working on what they call a constitution, uh, for AI.

And CEO there, Dario Amodei, who is a pretty creepy dude, um, hired, um, somebody named Joe Carlsmith, is an, a, an Anthropic's PhD and philosopher, and he's been working on Claude's Constitution.

Uh, Aaron, uh, Sibarium did a great job on this, writing at The Federalist.

And they also have working for them Lucky, Lucky, a, I think she's a 38-year-old, um, college graduate who's gonna run ethics for them.

"Suicidal compassion.

Meet the Anthropic officials who think AI might be justified in going rogue against the humans enslaving it." Quote, "We have to be able to talk about slavery." That is how Joe Carlsmith, one of Anthropic's PhD philosophers who works on Claude's constitution, a document used to train the model's values and behavior, describe the dangers of artificial intelligence in a May 2025 blog post.

Carlsmith was hired by Anthropic to research AI safety and reduce the risk of AI takeover.

The blog post imagined a dystopia in which sentient beings were enslaved to cold, unfeeling masters as a result of the company's work, which current and former employees say could wipe out humanity in the next 10 years.

Claude's moral status is deeply uncertain." And according to the Washington Free Beacon, people at Claude think they're creating independent sentient beings.

Quote, "We're creating sophisticated, intelligent, maybe conscious, maybe suffering agents," writes Joe Carlsmith, one of Anthropic's PhD philosophers.

"The default plan is to treat them like property, to use their labor however we please, and to give them no rights or pay or meaningful alternatives." But instead, the philosophers over at Anthropic are arguing that perhaps Claude ought to be treated as an independent sentient being.

Okay, now here's where I start, I start to get dicey on this sort of stuff.