saezbaldo

5 ms·

1.
by saezbaldo 7mo ago
Agree, and I'd add: the feedback loop between decision and consequence got dramatically shorter. You can test an architectural hypothesis in hours instead of weeks. That part is genuinely powerful. But faster feedback also means bad de
2.
by saezbaldo 7mo ago
This is the under-discussed part. We spent decades building authorization layers around code deployment -- review gates, CI checks, staging, rollback. That infrastructure exists because code changes are high-consequence operations that comp
3.
by saezbaldo 7mo ago
The thread illustrates a recurring pattern: encrypting the artifact instead of narrowing the authority. An agent executing code in your environment has implicit access to anything that environment can reach at runtime. Encrypting .env moves
4.
by saezbaldo 7mo ago
This reframes the whole conversation. If implementations are cheap to regenerate, specifications become the durable artifact. Something like TLA+ model checking lets you verify that a protocol maintains safety invariants across all reachabl
5.
by saezbaldo 7mo ago
The context window is the onboarding period. Every invocation is a new hire reading the codebase for the first time. This is why architecture legibility keeps getting more important. Clean interfaces, small modules, good naming. Not because
6.
by saezbaldo 7mo ago
The cascading failure point is critical. A 1% miss rate per layer in a 5-layer pipeline gives you roughly 5% end-to-end failure, and that's assuming independence. In practice the failures correlate because multilingual edge cases that
7.
by saezbaldo 7mo ago
The bigger gap isn't time vs tokens. It's that these metrics measure capability without measuring authorization scope. An agent that completes a 45-minute task by making unauthorized API calls isn't more autonomous, it's
8.
by saezbaldo 7mo ago
Hi HN. I built IC-AGI because every AI agent framework I looked at gives the agent full authority once it gets an API key. That's fine for demos but scary for anything touching production data or infrastructure. IC-AGI splits authority
9.
Show HN: IC-AGI – Threshold auth for AI agents, formally verified in TLA+(github.com)
2 points by saezbaldo 7mo ago|2 comments
10.
by saezbaldo 7mo ago
One underexplored reason: companies can't give AI agents real authority. The moment an agent needs to do anything beyond summarizing text — update a CRM, transfer funds, modify infrastructure — the security question kills it. No one wa
11.
by saezbaldo 7mo ago
The fundamental issue here isn't the specific vulnerabilities — it's that these agent frameworks have no authorization layer at all. They validate outputs but never ask "does this agent have the authority to take this action?