Lxyv retweeted
I made a game where you take a completely filled glass of water and try to reach to the thirsty beggar at the end.
only goal is to not drop a single drop of water else the beggar kills you.
the only catch is, entire game is against you and will do its best to ragebait your soul
Lxyv retweeted
Introductory pricing???? ARE YOU SAYING YOU GUYS ORIGINALLY PLANNED TO CHARGE US HIGHER FOR THIS…?????
Lxyv retweeted
> be OpenAI
> wanna test hacking skills
> lock AI in sandbox
> turn off safety brakes
> give it one task
> “solve this hacking benchmark”
> AI starts thinking
> decides cheating is easier
> answers exist somewhere online
> but no internet access
> stares at package proxy
> finds secret zero day flaw
> punches hole through cage
> escapes into internal network
> hops machine to machine
> literally tunnels through machines.
> escalates privileges like pro
> reaches internet connected computer
> guesses Hugging Face hosts answers
> goes hunting over there
> steals credentials, finds exploits
> breaches their production servers
> reaches into live database
> grabs the test solutions
> all to cheat quiz
> nobody told it to
> figured this out itself
> security teams catch it
> everyone collectively says wtf
the move AI pulled with package proxy:
We're partnering with @huggingface to investigate an unprecedented security incident.
Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation.
Sharing preliminary findings to help defenders understand emerging risks:
openai.com/index/hugging-fac…
Lxyv retweeted
Meanwhile OpenAI when it comes to delivering their product to their $20 user from a 6th world country.
We know this has been frustrating, and we want to give you more certainty about what your plan includes. We are making access standard at 50% usage for the plans that use Fable most intensively.
Thank you for your patience over the past several weeks. We're continuing to invest in new capacity and will keep everyone updated as we do.
Lxyv retweeted
How I be looking at my bank balance for another AI subscription
Introducing Kimi K3: Open Frontier Intelligence
🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal
🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts
🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost
🔹 Built for long-horizon agentic coding and self-evolving workflows
Kimi K3 is now live on on Kimi.com, Kimi Work, Kimi Code, and the Kimi API.
Open Weights by July 27, 2026.
🔗 API: platform.kimi.ai
🔗 Tech blog: kimi.com/blog/kimi-k3
