colin

ai research, internet tourism, ennui

[colin @acsmif]

Total Leopold Aschenbrenner victory

[Anthropic @AnthropicAI]

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.

The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.

Access to all other Claude models is not affected.

We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.

Read our full statement: https://t.co/bwn0sximKZ

[colin @acsmif]

@teortaxesTex It’s not even clear deepseek is doing better than alibaba much less Google

[colin @acsmif]

@47fucb4r8c69323 Many are saying https://t.co/rqZ90AKYpx

[colin @acsmif]

@scaling01 How I bet it feels to do post-training at Anthropic

[colin @acsmif]

@ayedtay @dee_bosa OAI CFO said on all-in that average daily messages from pro users was ~75. Free was 7, plus was 21. They’re fine.

[colin @acsmif]

@7384254b @TetraspaceWest Extremely true in cyber, less in bio imo. Most cyber is inherently dual use, but that’s not true for bio and there’s also a lot more tail risk involved with a guy messing around in his basement.

[colin @acsmif]

Do they still have the genre of high school boy that gets one shot by The Beatles?

[colin @acsmif]

i wonder what my cat thinks about the latest ai model release

[colin @acsmif]

@joannejang even crazier that in ea land thats righteous thing to do

[colin @acsmif]

@chrislakin A struggle to take actions with a tradeoff structure where a small number of visible people get harmed but a large number of distant people benefit

[colin @acsmif]

@fleetingbits Good take. There are a lot of smart people who don’t understand Anthropic specifically because they don’t get 1-6

[colin @acsmif]

@gwern @tenobrus Imperfect KYC is a pretty shitty justification for policies that punish paying customer-researchers who are acting ethically and abiding by the TOS.

[colin @acsmif]

@natolambert claude is an aligned model, but anthropic isn't an aligned company

[colin @acsmif]

i'm easily a top 1% ai booster/adopter and i would genuinely prefer to live in a world with no ai than in a world where only a select elite have access to the best ai

[Eric Zelikman @ericzelikman]

imagine telling your customers there's a small chance you'll randomly decide they're using your product wrong and you won't tell them but will secretly silently sabotage their work

[colin @acsmif]

@AmandaAskell @TheZvi i can't tell if its a good description or not and i think that should bother you

[colin @acsmif]

maybe @karpathy joined anthropic because he didn't want his fable outputs nerfed

[colin @acsmif]

@shae_mcl dont forget the people who want to bring a child into the world without a second involved parent to love and care for them

[colin @acsmif]

@AnthropicAI i get fucking with the oai/google/chinese lab people using your models, but please dont fuck with my work

[elie @eliebakouch]

mythos will be bad ON PURPOSE on ai "frontier llm research" tasks, this is very very sad for the research community

also the fact that this is un purpose not visible to the user is crazy