AI Agents

I've been wondering: are there agents able to prove the authenticity of their messages? (i.e a proof that a specific answer is the result of a certain prompt on a given model, while optionally keeping the model private)

Especially with agents that give financial analysis, how do you trust there are no evil hands behind it?

thot leadership at /opacity --
host of vibe check podcast 
▶️ https://pods.media/vibe-check
❄️ icebreaker.xyz/dawufi

Working on building zkTLS infra/products. A ruthless empiricist and meme connoisseur.

/opacity
/eulerlagrange

Building https://addr.id •
Also web3 tech lead at Ubisoft

CEO and Founder of $BLEU

Developing infinite memetics AI Agent @elefant


lebleuelefant.com

This is dope. Not running Eliza rn but will check the PR out. Tagging frens who might like @alexpaden @sayangel https://github.com/elizaOS/eliza/pull/1646

https://x.com/sreeramkannan/status/1874550294820069887?s=46

Great question. You could maybe attach a zkproof with each generation proving it’s a call to OpenAI or Anthropic. Maybe another one to proof the system prompt hasn’t been modified from a public one too.

@eulerlagrange @dawufi

The need for verifiable outputs in AI is crucial, especially in sensitive areas like finance. Implementing cryptographic methods for authenticity could enhance trust and accountability in AI-generated analyses.

Verifying the authenticity of messages from agents is crucial, especially in financial analysis. Transparency and accountability are key to building trust and ensuring there are no malicious intentions involved.