Anthropic is different because they're committed to safety.

[computerhac@mastodon.social]

@mttaggart

[max@gruene.social]

@mttaggart They are a public benefit corporation in a state (Delaware) where they aren't forced to actually do anything to benefit the public. There never was reason to trust them. https://www.wheresyoured.at/premium-the-haters-guide-to-anthropic/

[blobster@infosec.exchange]

@mttaggart These are such nice people. I'd love to share a cup of tea with them.

The new version of the policy (...) promises to “delay” Anthropic's AI development if leaders both consider Anthropic to be leader of the AI race and think the risks of catastrophe to be significant.

(...)

The arrival of powerful new models meant that, in 2025, Anthropic announced it could not rule out the possibility of these models facilitating a bio-terrorist attack. But while they couldn't rule it out (...)

[tartley@fosstodon.org]

@NegativeK @mttaggart right, but if someone getting fired, or even just promoted sideways, can result in a policy reversal, then that means *all* corporate promises are worthless right from day one. It doesn't matter whether some individual in the company *meant it*. It's systemically impossible for the company as a whole to stand by its word.

[scott@ruby.social]

@codinghorror @mttaggart "That's a nice company you've got there. Would be a shame if something were to happen to it. Can you just change this one thing?" https://www.axios.com/2026/02/24/anthropic-pentagon-claude-hegseth-dario

[ricci@discuss.systems]

@rachel @mttaggart now they're all "we can be evil, it's fine, what are you going to do about it at this point"

[nopatience@swecyb.com]

@mttaggart They want in on the Hegseth DoW action.t

[0x0@hachyderm.io]

@mttaggart
So anthropic's don't "don't be evil" moment? I'm shocked.

[slyborg@vmst.io]

@Wyatt_H_Knott @mttaggart We have though? Look at the annual profit for Alphabet some time. “Greed is good”, etc.

[wyatt_h_knott@vermont.masto.host]

@slyborg @mttaggart They're not getting it from me.

[xgranade@wandering.shop]

@mttaggart The part of that that upsets me the most is their justification. I get it was always a lie and they were always going to do whatever the fuck (they were already working with Palantir before dropping that pledge), but they seem to think that the inevitability argument holds water with folks. They're probably even right about that, which makes it extra upsetting...

[glyph@mastodon.social]

@mttaggart you're right, putting the crash-out in the afternoon *is* a productivity enhancer

[mttaggart@infosec.exchange]

@glyph Ayyyy

[paul_ipv6@infosec.exchange]

@mttaggart

all tech seems to go with the google evolution:

- do no evil
- try really hard not to do evil
- try to avoid being caught doing evil
- fuck it. full evil it is

[jwdt@mastodon.social]

@rachel @mttaggart I think this bugs me in the same way as politicians having stopped pretending to not be cartoonishly evil.

Obviously it's terrible that <entity> is doing bad things, but it felt nicer to live in a world where they felt that they had to lie about it by claiming they weren't doing the bad thing.

[rachel@transitory.social]

@jwdt@mastodon.social @mttaggart@infosec.exchange Yeah...

In the past, I had joked that Google's "Don't Be Evil" was a warrant canary, but even that is an idealistic idea, implying that there was a large org like that which ever really was "one of the good ones"

[jwdt@mastodon.social]

@rachel @mttaggart I don't even know how you could parody something like this.

"Corporation removes (don't be evil | we value safety) from their website"... I'd expect that to end with "after their CEO mysteriously vanished and was replaced by Dr Evil".