This makes me want to see China and open models succeed more than anything :)
replyDon't worry, we will succeed :)
replyCan we get a Qwen3.7-122B, please? Thank you.
replyIt's bad that Anthropic can determine what this means. If you're building a modern app you're likely training your own embedding models and now anthropic can just silently sabotage your training pipelines?
reply>We estimate they will impact ~0.03% of traffic, concentrated in fewer than 0.1% of organizations
replyAt the scale of API requests that Anthropic sees, I think the affected organization count might be substantial, and they might not be getting the full model capability that they're paying top $$$ for.
Also, wonder how they arrived at that estimation.
One in 1000 organizations and one in 3000 requests is indeed a lot
replyLooks like Anthropic's definition of safety includes their own safety from competition.
replyAI vendor’s idea of safety has always been safety for the interests of the AI vendor in question. This is not a new development, though this may help more people realize it.
replyAI-generated competition for thee, not for me
replyIt's always been about the safety of their valuation.
replyOnly since Claude 3. So a bit over two years now
replyHow do they detect whether an experiment being done on a smaller model is used to improve a competing frontier model, or just an innocuous hobbyist LLM experiment?
replyGiven how well the cybersecurity safeguards work, they probably don't.
replyinfering the surroundings, like everything else. they will probably look at which company is your email, and if you wrote "better than claude" on the readme.md
replythis is LLM, it's not like a science or something.
Ah, so this is why raw Mythos was too "dangerous" to realease..
replyA million AI researcher voices at big tech companies suddenly cried out in terror and were suddenly silenced
reply3 months before asking for what to eat before a linear algebra exam trips the machine learning topic ban is my guess. I got flagged immediately asking why my JEPA thing breaks weird.
replyMeaningless and easily bypassable. Will actually try coding up a tensor library with it, see if it sabotages anything.
replyThey said in their terms and conditions they will silently sabotage you if you do this.
replyeasily ?
replydeleted
replyThis is pretty bullshit, now you have no idea if your output is getting silently nerfed.
replyIt's afraid!
replyYeesh. Anthropic's paranoia about China is starting to get pathological.
reply