‹ BackHN Continuity

Thread

Astra for Law

589 points · 689 comments · vertigoruntime

  1. cbg0 · · focus · HN ↗
    No word on model hallucinations in the blog post.

    <a href="https:&#x2F;&#x2F;artificialanalysis.ai&#x2F;models&#x2F;gpt-6-astra?omniscience=omniscience-hallucination-rate" rel="nofollow">https:&#x2F;&#x2F;artificialanalysis.ai&#x2F;models&#x2F;gpt-6-astra?omniscience...

    1. jrflo · · focus · HN ↗
      Does hallucination matter for this application? We&#x27;ve moved beyond raw recall being that important, it seems like for law specifically all relevant facts will be cited and checked easily by humans.
      1. Implicated · · focus · HN ↗
        &gt; it seems like for law specifically all relevant facts will be cited and checked easily by humans.

        Don&#x27;t be too sure about that. [0]

        0: <a href="https:&#x2F;&#x2F;www.damiencharlotin.com&#x2F;hallucinations&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.damiencharlotin.com&#x2F;hallucinations&#x2F;

      2. jakevoytko · · focus · HN ↗
        &gt; all relevant facts will be cited and checked easily by humans

        I&#x27;ve talked to a lawyer about how they handle this. They do indeed double-check everything, since it&#x27;d be embarrassing (or worse) to send hallucinated statements to opposing council or to the court. They still find the assembly a huge time saver

        But based on stories in the news on the subject, not everyone has this same level of diligence

      3. samtp · · focus · HN ↗
        Sure but just like generating 100x more code, someone has to review it. So you are wasting everyone in court&#x27;s time (defendants, prosecutors, judges, staff) by making them parse through what is quite often a bunch of hallucinated slop. Time that could be much better spent on parties who prepared and reviews their own arguments.
      4. jordanpg · · focus · HN ↗
        The lawyers I know are very fixated on the problem of hallucinated case citations which is amusing to me as a onetime programmer, since case citations have a well-defined syntax and would be relatively easy to check programmatically.
        1. massysett · · focus · HN ↗
          Which makes it all the more bizarre that LLMs have this problem. Claude Code runs the code it generates through a compiler, why can’t an LLM run its product through a cite checker? I’ve seen LLMs fabricate citations.
          1. heaney-555 · · focus · HN ↗
            A modern frontier model with an agentic harness will check its citations if the system instructions or prompt tell it to.

            You can get an effectively-zero hallucination rate with the right setup already.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.