Also, mass surveillance. Not surveillance itself. And fully autonomous weapons.
Don’t get distracted by the birdy folks, Anthropic is not your friend, or some great protector of the American people. They were already deeply embedded in the US Government as their product was the only one certified for use with classified documents.
They weren’t standing up for us, they were splitting hairs on exactly how far they’d openly go.
I’ve also seen statements that Anthropic’s stance against fully autonomous weapons was simply due to results not yet being as consistent as they were comfortable putting their name on, not due to any opposition towards use in/with weaponry.
OpenAI also claims to have the same limitations. So someone’s lying.
Hackworth@piefed.ca 6 days ago
Amodei said in an interview that the DoW altered their contract to appear to compromise, so that it looked like they were agreeing to those use limits. But that legalese accompanying the updates rendered that text pointless. Basically, “We won’t use Claude for mass domestic surveillance and full automated killing, unless we really want to.” My guess is OpenAI signed the exact same contract and just pretended not to understand the toothlessness of the guardrails.
XLE@piefed.social 6 days ago
“Guardrail” and “toothless” are basically synonymous, based on the pile of evidence that these multi-billion-dollar tech companies have been helping people kill themselves and hide the evidence.
mgenehoffman@sh.itjust.works 6 days ago
Or, even more ironically, maybe they used ChatGPT to analyze the changes and it missed it. This would tickle me to some extent, but also solidify the terror of such a system being used to make life altering decisions.