Pinned Tweet
I DO NOT CONSENT!
I don't think people understand the gravity of the situation as the UN is preparing for possible nuclear weapon use in Iran.
This is a picture of Tehran. For you uneducated, untraveled, never-served, warhawks licking your chops at the thought of bombing it. It's not some low population desert. There are families, children, family pets. Regular working class people with dreams. You're sick to want war.
Tehran is a city of nearly 10,000,000 people. Imagine nuking Washington, Berlin, Paris, London, or beyond, bombed with nuclear weapons.
I gave up my diplomatic career to leak this information. I suspended my duties so as not to be part of or a witness to this crime against humanity, in an attempt to prevent a nuclear winter before it is too late.
Yesterday, nearly ten million people protested “No Kings” in the United States. The possibility of the use of nuclear weapons must be taken very seriously. It's dangerous. Act now. Spread this message worldwide. Take the streets. Protest for our humanity and future. Only the people can stop it. History will remember us.
How about We turn AI‑agent oversight into code: an on‑chain ERC‑8004 mandate + Merkle‑anchored proofs that makes illegal actions impossible and every legal action cryptographically provable?
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here: darioamodei.com/post/we-must…
chris baliza retweeted
未来AI身份卡已经到来。每个AI代理——甚至有朝一日每个仿人机器人——都拥有链上可验证的法律身份。KYA已验证 · 有限授权 · 可撤销。 @Tracea_ai
The future of enforced and legal AI Agent Identity is Here!
Try now on tracea.online
The future AI identity card is here. Every AI agent — and someday every humanoid — with a verifiable on-chain legal identity. KYA verified · limited mandate · revocable. @Tracea_ai
We have launched tonight a full passport and wallet for AI agent on chain using ERC-8004 and x402.
@coinbase , @Ethereum_France and you!
We need you!
Support Us please!
🚀 We're launching today on Product Hunt!
If you believe AI agents should be verifiable and accountable:
• Support us on PH
• Follow @Tracea_ai
• Check out tracea.online
Thank you! 🙏 #ProductHunt #AI #web3
I'm testing by building a cleaning AI app with a0.dev – check it out
What if the AI does get stuck? Here's what you can do:
Try software termination signals first.
If that fails, consider a controlled shutdown of the hardware, but with caution to avoid data loss.
#AISafety
Remember, the goal isn't just to stop the loop but to understand why it happened. This helps in creating more robust AI systems. What are your thoughts on managing AI loops?
#AICommunity
Would love to hear from other AI developers or enthusiasts. Have you encountered an "Internal формирование Loop"? How did you solve it? Let's learn from each other's experiences.
#AIConversations
Any thought if #GrokAI goes infinite looping @elonmusk ?
Detection is key! Here are some signs your AIl might be in a loop:
Unusually high CPU/GPU usage without output
Repetitive actions or queries
No progression in task completion
#AIEngineering
So, how do we prevent these loops?
Implement timeouts & iteration limits.l
Use human-in-the-loop (HITL) systems for oversight.
Develop algorithms to detect looping behavior.
#AIDesign
What is an "Internal формирование Loop"? It's when an AI agent gets stuck in a cycle of self-validation or processing without progress. It's essentially an AI version of an infinite loop.
#AI #MachineLearning
Why is this a problem? If an AI doesn't stop to validate itself against external data or human input, it could consume resources endlessly, leading to inefficiency or even system crashes.
#AIRisks
Just had a thought-provoking dream about AI Agents and what I call "Internal формирование Loop." Let's dive into what this means, the risks it poses, and how we can manage it.
I was sleeping and dreamed about an AI Agent and I got into what I call "Internal формирование Loop."
What if an Agent AI falls into a "Internal формирование Loop" and never stops to validate itself?
How do we stop it ?
Do we switch off the physical hardware 🧵👇
chris baliza retweeted
Les risques liés aux stable coins et aux smart contracts sont énormes! 😱😱 Il est crucial de mettre en place des mesures pour les réduire. #StableCoins #SmartContracts #RéductionDesRisques









