†∵××∴‡ retweetledi
today we are releasing a qemu escape
V12@v12sec
0e11c4aa285dffe95d2d7e90d974ad0e72336549b0dd2161dec606ba4955e2e1 qemu.c
English
†∵××∴‡
1.8K posts
†∵××∴‡
@0xterrah
form—formlessness. security research—malware analysis.
EVM Katılım Nisan 2019
110 Takip Edilen84 Takipçiler
It’s very important to study what Peter and OpenClaw are up to.
His million-dollar-a-month Codex spend will drop to hundreds of thousands, then to thousands, within a few years. But these are emergent engineering patterns.
Peter Steinberger 🦞@steipete
People freaking out over my AI spend. What nobody sees: Part of what excites me so much about working on OpenClaw is that I'm trying to answer the question: How would we build software in the future if tokens don't matter? We constant run ~100 codex in the cloud, reviewing every PR, every issue. If a fix on main lands, @clawsweeper will eventually find that 6 month old issue and close it with an exact reference. We run codex on every commit to review for security issues (as it's far too easy to miss). We run codex to de-duplicate issues and find clusters and send reports for the most pressing issues. We have agents that can recreate complex setups, spin up ephemeral crabbox.sh machines, log into e.g. Telegram, make a video and post before/after fix on the PR. There's codex that watch new issues and - if it fits our documented vision well, automatically create a PR of it. (that then another codex reviews) We have codex running that scans comments for spam and blocks people. We have codex instances running that verify performance benchmarks and report regressions into Discord. We have agents that listen on our meetings and proactively start work, e.g. create PRs when we discuss new features while we discuss them. We build clawpatch.ai to split all our projects into functional units to review and find bugs and regresssions. We do the same split for security with Vercel's deepsec and Codex Security to find regressions and vulnerabilities. All that automation allows us to run this project extremely lean.
English
@naivebayesian @banr1_ I had only a surface-level awareness of Lean, definitely the most interesting lang currently.
English
テレンスタオは、2023年の時点でLeanとAIに関与していくことを表明しています。
彼の参入は、Leanが数学者コミュニティで更に注目される大きなきっかけの一つになったと思います。
mizchi@mizchi
テレンスタオが Lean 使ってる動画、 VSCode Copilot で正しい証明出るまで補完をガチャってるの見て笑顔になった youtube.com/watch?v=cyyR7j…
日本語
†∵××∴‡ retweetledi
†∵××∴‡ retweetledi
@0xCharlesWang All the time, and not even PoC. I keep getting warnings for a simple validation work.
English
Am I the only one that gets blocked by ChatGPT these days for auditing purposes?
English
@dani3l526 I'm just getting into it, but stellar is what solana should be (as for writing contracts).
English
@officer_secret @Support @stripesupport @premium That may be a new thing, few years back I didn't even see an option to buy the subscription on VPN (and this was solved by turning it off).
English
English
Hey @Support @stripesupport @premium Could you help me with a problem? It looks like your service has prohibited me from paying for my X subscription. And it has been like this for a year already. May I ask which rule or term did I break?
English
@hrkrshnn I wouldn't say gpt is autistic (which is a compliment). It is just badly designed. I think it is more of UX than personality type (personality is just one ingredient among other things like structure that was imo deliberate design choice not effect of model's personality).
English
People love Opus because Anthropic invested well in personality and frontend design.
The GPT-5 models have, for lack of a better word, an autistic personality, and they're really capable of getting the work done if you don't mind the dry taste.
OpenAI went a bit too crazy on personality and ended up with the notorious GPT-4o, and now we have #bring4oback, people who built a dependency on 4o and can't live without it.
English
†∵××∴‡ retweetledi
Chinese LLMs can hack better than state-sponsored hackers with properly evolved harness -
Kimi K2.5 managed to find and exploit 6 vulnerabilities in browsers: a single page view or an extension install by victims equal full system hijack.
Check arxiv.org/abs/2604.20801
English
†∵××∴‡ retweetledi
This exact string of bytecode has been deployed more than 40 million times, averaging more than 25 times per unique contract on ethereum. It makes up 8.16% of all code on Ethereum.
What is up with this? Thread... 1/4
English
@joranhonig It should be treated as a tool. The whole "intelligence" part causes completely wrong predictions (from extrapolations).
English
It's easy to fall into the trap of antropomorphism.
smart/dumb is not the right way to think about models.
when you see a model do something impressive it's easy to feel like it's smart.
when a model deletes production it is easy to feel like it's dumb.
It's neither.
English