semisol on Nostr: It is weird to see that people trust AI to evaluate the security of whatever HWW they ...
It is weird to see that people trust AI to evaluate the security of whatever HWW they use.
These models cannot form independent opinion, and only repeat what people already said.
From my testing, they get an egregious number of things extremely wrong.
But I guess these models are trained to *sound* like an expert in everything, so…
Published at
2026-08-01 20:37:29 CESTEvent JSON
{
"id": "b3cb42c08dab8c2fdd90333e651c4269ad915e7dc69865a1d6d8157cb5c39a97",
"pubkey": "52b4a076bcbbbdc3a1aefa3735816cf74993b1b8db202b01c883c58be7fad8bd",
"created_at": 1785609449,
"kind": 1,
"tags": [
[
"client",
"Damus"
]
],
"content": "It is weird to see that people trust AI to evaluate the security of whatever HWW they use.\n\nThese models cannot form independent opinion, and only repeat what people already said.\n\nFrom my testing, they get an egregious number of things extremely wrong.\n\nBut I guess these models are trained to *sound* like an expert in everything, so…",
"sig": "9897e5f751269b9b169dcb1bb2d80f1b008d5c29154649f5238db6a770cc92b8b1d25bb2a5a01289a154442c6fced203b598f0446972a49ac3bfe5141f74a8f8"
}
