{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiaguzibficwk76u6zj73pbpyvfnirlovxoxg6mzmc5mqwcoxwgfii",
    "uri": "at://did:plc:pi6woz4d47bkuws673w2il2r/app.bsky.feed.post/3mh4cgx4q2u22"
  },
  "path": "/t/yet-another-opinion-on-llms-hasufells-blog/13775?page=2#post_27",
  "publishedAt": "2026-03-15T08:50:16.000Z",
  "site": "https://discourse.haskell.org",
  "tags": [
    "And we had some high profile examples"
  ],
  "textContent": "Swordlash:\n\n> I honestly never had if hallucinating, but maybe I didn’t notice?\n\nI find it hard to imagine how you used those tools without it hallucinating at some point. Maybe you mentally classified those issues as it “just being wrong” or something?\n\nBecause for me it happens pretty often. And yes, with “reasoning” often they will correct themselves before presenting you a final solution. But sometimes they won’t. And sure we can say any particular hallucination could have been prevented if there were more context, if the prompt had been better, or some other reason. But the whole issue is that one can’t know what input will lead to hallucinations beforehand.\n\nWhich in turn means any output should ideally be considered untrustworthy. But that is just not a cognitive load that people can maintain in the long run. I had some experience in fields where safety and defects come with much higher costs and even there, long before AI, it was hard to ensure outputs are properly vetted and checked. And we had some high profile examples of such issues in aviation of all fields recently.\n\nIt’s clear to me in practice most companies will either accept a higher defect rate in exchange for higher productivity or they end up taking additional measures (more types, more tests, more invariants, more specific context) in order to minimize the risk of defects which eats into the increased productivity. And for the later situation it’s very unclear to me how much of a win those tools really are.",
  "title": "Yet another opinion on LLMs · Hasufell's blog"
}