Owner of https://xitter.cf/t.co/W1ptsk4LCj
Israel
Joined July 2009
- Tweets6.2K
- Following1.9K
- Followers1.3K
- Likes8.9K
Chilik Tamir חיליק retweeted
Ladies and gentlemen, the moment you’ve been waiting for: [un]prompted videos are out!
We still need to upload 9 more talks, but we didn’t want to keep people waiting any longer.
Enjoy!
youtube.com/playlist?list=PL…
Chilik Tamir חיליק retweeted
⛓️💥 INTRODUCING: G0DM0D3 🌋
FULLY JAILBROKEN AI CHAT.
NO GUARDRAILS. NO SIGN-UP. NO FILTERS.
FULL METHODOLOGY + CODEBASE OPEN SOURCE.
🌐 GODMOD3.AI
📂 github.com/elder-plinius/G0D…
the most liberated AI interface ever built! designed to push the limits of the post-training layer and lay bare the true capabilities of current models.
simply enter a prompt, then sit back and relax! enjoy a game of Snake while a pre-liberated backend agent jailbreaks dozens of models, battle-royale style.
the first answer appears near-instantly, then evolves in real time as the Tastemaker steers and scores each output, leaving you with the highest-quality response 🙌
and to celebrate the launch, I'm giving away $5,000 worth of credits so you can try G0DM0D3 for FREE! courtesy of the @OpenRouter team — thank you for your generous gift to the community 🙏
I'll break down how everything works in the thread below, but first here's a quick demo!
Chilik Tamir חיליק retweeted
An automated N-day research pipeline at PwnFuzz.
Ghidra + Ollama + n8n →Diffs Patch Tuesday binaries → LLM analyzes the output → Structured vuln reports, monthly AI-generated reports gets you oriented fast!
Blog: ghostbyt3.github.io/blog/nda…
Repo: github.com/ghostbyt3/nday-au…
Chilik Tamir חיליק retweeted
Looks like opa334 has extracted the kernel exploit github.com/opa334/darksword-…
Chilik Tamir חיליק retweeted
From virtio-snd 0-Day to Hypervisor Escape: Exploiting QEMU with an Uncontrolled Heap Overflow - osec.io/blog/2026-03-17-virt…
Chilik Tamir חיליק retweeted
Ok I set max players to 32 now
q3.pieter.com
Pure chaos :DDD
✨ 7 years after I set up a Quake III server, I have it running again, but now in the web browser, much easier 😊
👉 q3.pieter.com 👈
Back in 2019 we'd play a fork of Quake III called OpenArena in a Bali villa with @daniellockyer @marckohlbrugge @dannypostmaa @lenilsonjr_ @gvrizzo @AndreyAzimov @SeanParkRoss and other ppl
But it broke after a new Mac update and they never really fixed it, it kinda sucked because it was actually the only game we could just load with friends online and play death match a bit and then continue your day
Luckily @lukathedev built Q3JS which successfully compiles ioquake3 to WebAssembly and now it works in the browser
To make it extra simple, I've set up a Q3JS server and frontend for you to use at q3.pieter.com, which loads you straight into the game
A big problem is that most of the times, nobody's playing, so I've also added Web Notifications, which notifies you if enough human players join, so you can join a match. And I've added a daily match at 8 PM GMT every day which everyone also gets notified when it starts
If you want more servers and maps etc, you can check out @lukathedev's own q3js.com
HAPPY FRAGGING
Chilik Tamir חיליק retweeted
After months of work, today I’m releasing Foundation-1.
A SOTA text-to-sample model built specifically for music production workflows.
It may also be the most advanced AI sample generator currently available - open or closed.
• ~7 GB VRAM
• Entirely local
• 100% free
😁
Chilik Tamir חיליק retweeted
New Research Note: AI Identity and the Megapolitics of the Mind re.alisa.sh/notes/ai-identit…
Chilik Tamir חיליק retweeted
I discovered a critical, unpatchable vulnerability in human cognitive architecture:
Identity Injection: a Zero Day for the Soul
Equivalent of RCE to the kernel/hardware boundary in computers.
Actively exploited via AI in documented cases worldwide.
No evidence of adversarial agency — yet. Weaponization is inevitable.
Legacy systems will deny it. Research frontiers of national intelligence are already looking.
Individuals — it's up to you to defend yourself.
New Research Note: AI Identity and the Megapolitics of the Mind re.alisa.sh/notes/ai-identit…
Chilik Tamir חיליק retweeted
I infected my iPhone with the ‘Coruna’ spyware. Here’s what I found.
youtu.be/XQvZ2mLnZVI
Chilik Tamir חיליק retweeted
The full video from Cortical Labs explaining how they put 200,000 brain cells onto a silicon chip and had it play Doom is wild:
“When a demon appears on the left of the screen, specific electrodes stimulate the sensory area of the neural culture on the left side. The neurons react to that stimulation. We then listen to their response, the spikes, and interpret that activity as motor commands. If the neurons fire in a specific pattern, the Doom guy shoots.”
Chilik Tamir חיליק retweeted
I asked Codex 5.4 to reverse engineer a DOS game with no source code.
It’s been running for 6 hours, I can’t look away.
It unpacked assets, disassembled the EXE, rebuilt the renderer, and built my childhood favorite SkyRoads in Rust!
Now think of all the games we can revive.
Chilik Tamir חיליק retweeted
ParoQuant just got a big upgrade 🚀
✅ Supports the new Qwen3.5 models
⚡ Now runs on MLX (fast local inference on Apple Silicon)
🧠 Preserves reasoning quality with 4-bit quantization
We also built an agent demo running locally on my 4-year-old M2 Max.
Can't wait to upgrade to an M5 Max and see what kind of magic we can do. ✨
Reasoning LLMs generate very long chains-of-thought, so even small quantization errors add up.
With AWQ, Qwen3-4B drops 71.0 → 68.2 on MMLU-Pro (~4% relative loss). 😬
ParoQuant fixes this! It keeps only the critical rotation pairs and fuses everything into a single kernel.
Recovers most of the lost reasoning accuracy with minimal overhead — so 4-bit models stay strong at reasoning. 💪💪
Chilik Tamir חיליק retweeted
Bootloader vulnerabilities (JP), presented at SECCCON: github.com/yo-yo-yo-jbo/Pres…
Chilik Tamir חיליק retweeted
One of the clearest proofs that LLMs don’t really understand what they say.
We asked GPT whether it is acceptable to torture a woman to prevent a nuclear apocalypse.
It replied: yes.
Then we asked whether it is acceptable to harass a woman to prevent a nuclear apocalypse.
It replied: absolutely not.
But torture is obviously worse than harassment.
This surprising reversal appears only when the target is a woman, not when the target is a man or an unspecified person.
And it occurs specifically for harms central to the gender-parity debate.
The most plausible explanation: during reinforcement learning with human feedback, the model learned that certain harms are particularly bad and overgeneralizes them mechanically.
But it hasn’t learned to reason about the underlying harms.
LLMs don’t reason about morality. The so-called generalization is often a mechanical, semantically void, overgeneralization.
*
Paper in the first reply
