The company reportedly hired a rapid-response communications specialist.
Between Politico and Axios we get a pretty clear picture that this was another case of 'the vibes are off and when we say jump they ask for an explanation rather than how high, so let's blow everything up.'
Plus more whining about not getting Dario on the phone fast enough.
Many users criticized Anthropic's slow responses and lack of seriousness as inept governance by unqualified leaders like Dario.
@TheZvi This whole thing feels off, and I maintain that there's more to it than we know yet.
And honestly, the framing of Anthropic and Dario across multiple channels and talking heads and social media "influencers" with the exact same talking points is pretty telling.
@TheZvi DARIO: "I control the greatest weapon ever seen! The government cant control me!"
*gets controlled*
You: "Something seems off"
What the fuck are you people smoking?
@TheZvi I think at this point they may need a new CEO. Dario hasn’t handled these situations amazingly but more importantly this administration clearly hates the guy now and are willing to destroy American capitalism to punish him specifically.
@TheZvi if you say you are building a literal nuke and an economic chaos engine than yes -- you best be expecting the central gov to come knocking lol.
Gov may have been unreasonable but when Ant keeps marketing their models could bring an ELE then, under that premise, Ant’s response is beyond shocking. Either the models are an existential risk and Ant should be all hands on deck at the inkling of a problem or they aren’t and Ant needs to stop acting like they are. Can’t have it both ways.
@TheZvi US AI Policy... brought to you by a combination of Tony Soprano, The Joker, and 12 guys from the local truck driver's union.
@TheZvi The trouble is when you have no theory of mind about your opposition but they have theory of mind about you.
@TheZvi Your take on this seems genuinely strange. Do you think the government should wait until they are completely certain something poses a severe national security threat before acting?
@TheZvi If the U.S. government’s position is that no AI model can be released unless it is guaranteed to be immune to jailbreaks, then that effectively ends the AI industry. AI systems are inherently non-deterministic, which means no such guarantee is possible.
@shanboxfantasy @LyraInTheFlesh @TheZvi Because they don't generally lie (can you think of an example?) and this administration very routinely does with norm-defying boldness and frequency.
@TheZvi You didn’t mention that Amazon CEO was the person telling the USG that it was an urgent problem. Maybe Amazon is an investor in Anthropic, or something
Sure but you seem to have missed my point. I think Anthropic’s response is fine but it’s inconsistent with their own statements about their models. If you tell me you have created a nuclear bomb so you shield it and someone finds a way to get through that shield and get to the bomb even it’s just a radiation leak then yeah you need to call 911. Unless it’s not a nuclear bomb and then you need to stop saying it’s a nuclear bomb.
@YeOldStinkPot @TheZvi Last time the government tried to punish Anthropic it led to the greatest jump in sales the company had ever seen.
@TheZvi I'd be open to paying $300 for a subscription with the same rate limits but a "how high?" CEO
@TheZvi Hegseth’s tweet confirms this is all part of the payback for Anthropic refusing to bend the knee. Only path forward is a massive bribe from Dario to Trump Family Inc.
@LyraInTheFlesh @TheZvi I can’t understand why people take Anthropic at their word.
@hallmark_nick @TheZvi My car can definitely kill me or other people, but that doesn’t mean the moment I have an inkling something is wrong I need to pull over and call 911. Many problems with dangerous things are not closely tied to the most serious risks.
@TheZvi Cry
@TheZvi The right question is: why does USG have these issues with Anthropic but not OAI or Google?
@PJOPJOPJOPJO @TheZvi Your logic doesn’t hold up here dude. You are saying that inciting ongoing conflict with the U.S. gov is a good thing because one time it worked in their favor. I think we can disagree about how to avoid the conflict but I just fully reject that they benefit from it long term
