fabsalvadori

5 ms·

1.
by fabsalvadori 5d ago
Parmigiano Reggiano is a registered brand. It gives me goosebumps to read "parmesan" which is an unregulated translation or generic term used globally for very low quality imitations. Anyway, if you like cheese but you're int
2.
by fabsalvadori 23d ago
This is the line between an instruction and a control. If the agent can reinterpret, edit or relax the rule that constrains it, the rule isn't actually enforcing anything... it's just part of the prompt. I think the useful split i
3.
by fabsalvadori 26d ago
There is a second-order effect I see, here: when generating work becomes dramatically cheaper, failed work also becomes dramatically more common. With coding agents, cheap intelligence doesn't just mean producing the same software for
4.
by fabsalvadori 27d ago
There is a useful engineering consequence here beyond terminology. If intermediate tokens are not a faithful representation of the computation, then they are a pretty bad audit artifact too. We probably shouldn't be trying to make the
5.
by fabsalvadori 27d ago
Interesting results, but the fix is at the wrong level. If the model can access something, telling it in the prompt not to use it is not much of a safeguard. The strongest evidence is in the results: when one way of cheating was discouraged
6.
by fabsalvadori 1mo ago
I see little conversation about AI powering robots' core and the fact that we are going to have like a billion robots by the end of this decade. Yes, sorry if trust is low.
7.
AI agents are a confused deputy with the keys to your kingdom(stackoverflow.blog)
3 points by fabsalvadori 3mo ago|0 comments
8.
Fail-closed evidence for LLM tool calls (SHA-256 and MCP)(github.com)
1 points by fabsalvadori 8mo ago|0 comments
9.
Show HN: Pic-Standard – Open Protocol for Agentic AI Safety(github.com)
1 points by fabsalvadori 8mo ago|0 comments