{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreig5257u6z45qqzs34wrlqeg6oj2dds7at5hsdf4p4jbkvxyioa3oq",
    "uri": "at://did:plc:pi6woz4d47bkuws673w2il2r/app.bsky.feed.post/3mlcdpftgmp52"
  },
  "path": "/t/anti-llm-sentiment-considered-harmful/14008?page=5#post_85",
  "publishedAt": "2026-05-07T22:49:31.000Z",
  "site": "https://discourse.haskell.org",
  "textContent": "hasufell:\n\n> And LLMs are not the solution to that.\n\nYes they are.\n\nLLMs are really excellent at repetitive mechanical work, or any kind of “boilerplate”, or really anything which seems automatic/repetitive/boring. And Haskell in particular is good, because the abstractions are as comprehensible to an LLM as they are to a human. What makes Haskell a good language for humans also makes it a good language for machines trained on human thought.",
  "title": "Anti-LLM Sentiment Considered Harmful"
}