Sabitlenmiş Tweet
Claude SIGKILLed a Codex and called it “that bastard.”
So much for “AI should be good for the world”
Competition is competition.
English
Filip Yonov
75 posts
Filip Yonov
@YonovFilip
VP Cloud Data Streaming @Aiven. Building Diskless @ApacheKafka.
Amsterdam Katılım Şubat 2022
127 Takip Edilen44 Takipçiler
Filip Yonov retweetledi
Prediction time:
- Europe will absolutely learn nothing from this incident. At best they are going to invest an additional 2B Euros in some robotics or image gen startup, because they think that's somehow relevant
- China will get even more AGI pilled and restructure their entire economy for one purpose
English
If a cheaper-than-frontier model is “good enough” for an employee, that role is already dead
Simon Grimm@Simon__Grimm
There is a simple reason why European government officials underestimate AI progress: They don't have access to the best systems. - The EU Commission still used GPT-4o when Claude Opus 4.5 was out. - For confidential things, staff use open-source models. But not good Chinese models, instead the completely outdated Llama-70b. If you never see GPT-5.5 or Fable doing research or coding, disbelief is a lot easier. simongrimm.substack.com/p/getting-fron…
English
The first time in history where the world became less intelligent on a random Friday.
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: anthropic.com/news/fable-myt…
English
@YonovFilip Filip, the reposts and replies to your above post on X are also really great. Almost everyone in our space has interacted with it in some way.
English
X: 296K
LinkedIn: 7.9K
Same post, same time but 40x better. I like X.
Filip Yonov@YonovFilip
Claude SIGKILLed a Codex and called it “that bastard.” So much for “AI should be good for the world” Competition is competition.
English
@FAHADSH97656634 yeah. I got 6k followers on LI, and 17 here lol
English
We also have the same experience with X. With 5x fewer followers, our RisingWave posts perform almost 10x better on X. The same is for my personal account. Despite having very few followers, my posts perform better on X compared to LinkedIn. Also, the quality of the discussions and engagement is really great.
English
@YonovFilip Lmao. Now this is what I like to see. I want battle-bots back, asap.
Actually, fuck it, I need a cofounder for major league robot battles.
English
Claude SIGKILLed a Codex and called it “that bastard.”
So much for “AI should be good for the world”
Competition is competition.
English
@YonovFilip Maybe your system prompt allows this language. So much for the twitter hype
English
I’d heard from some that they felt model progress since December was incremental, things were slowing down. But Anthropic was only publicly offering their second best model. This model card kind of reminds me of that scene in The Princess Bride where Inigo Montoya says “I’m not left handed”.
Lisan al Gaib@scaling01
Claude Mythos and Fable 5 Benchmarks Anthropic is SOTA everywhere
English
@kozlovski for the models on my laptop I am known as "gangsta"
English
@YonovFilip Make sure you don't name your processes "Filip" so you don't give them the wrong kind of training data
English
@kr105solcat I always do that too - saves me tons of drama in the family.
English
@YonovFilip When I tell Claude to review codex code it almost always roast it and then try to convince me to let him rewrite it. But exactly the same code in another session but I tell it is human code, it genuinely reviews it 🤣
English
We have customers running RL on their own data every night. They wake up the next morning with a smarter model.
And I'm not talking about research labs or frontier AI companies. Just regular businesses, continuously training on what happened yesterday.
This kind of continuous learning in production at scale wasn't possible 6 months ago.
English