8 comments

  • zOneLetter2 hours ago
    Anecdotally, we use an LLM note-taker at work for meetings. I had to intervene recently because our CIO was VERY angry at our vendor for something they promised to do and never did. He wasn&#x27;t at the meeting where the &quot;promise&quot; was made. I was. They never promised anything, and the discussion was significantly more nuanced than what the LLM wrote in the detailed summary.<p>In other cases, I have seen it miss the mark when the discussion is not very linear. For example, if I am going back and forth with the SOC team about their response to a recent alert&#x2F;incident. It&#x27;ll get the gist of it right, but if you&#x27;re relying on it for accuracy, holy hell does it miss the mark.<p>I can see the LLM take great notes for that initial nurse visit when you&#x27;re at the hospital: summarize your main issue, weight, height, recent changes, etc. I would not trust it when it comes to a detailed and technical back-and-forth with the doctor. I would think for compliance reasons hospitals would not want to alter the records and only go by transcripts, but what do I know...
    • fc417fc8023 minutes ago
      &gt; I would think for compliance reasons hospitals would not want to alter the records and only go by transcripts, but what do I know...<p>I&#x27;m puzzled by this as well. Why not just generate a transcript and be done with it? If it&#x27;s a particularly long transcript that&#x27;s being referenced repeatedly for whatever reason let the humans manually mark it up with a side by side summary when and where they feel the need. At least my experience is that usually these sort of interactions don&#x27;t have a lot of extraneous data that can be casually filtered out to begin with. The details tend to matter quite a lot!
    • Ferret74463 minutes ago
      Transcription works pretty well in my experience, and the transcripts should be treated as the ground truth in such cases.
  • rainsford1 hour ago
    I have generally moved from bearish to bullish on the future of current AI technology, but the continued inaccuracy with basic facts all while the models significantly improve continues to give me significant pause.<p>As an example, creating recipes with Claude Opus based on flavor profiles and preferences feels magical, right up until the point at which it can&#x27;t accurately convert between tablespoons and teaspoons. It&#x27;s like the point in the movie where a character is acting nearly right but something is a bit off and then it turns out they&#x27;re a zombie and going to try to eat your brain. This note taking example feels similar. It nearly works in some pretty impressive ways and then fails at the important details in a way that something able to do the things AI can allegedly do really shouldn&#x27;t.<p>It&#x27;s these failures that make me more and more convinced that while current generation AI can do some pretty cool things if you manage it right, we&#x27;re not actually on the right track to achieve real intelligence. The persistence of these incredibly basic failure modes even as models advance makes it fairly obvious that continued advancement isn&#x27;t going to actually address those problems.
    • Brian_K_White1 hour ago
      I hate to help provide possible soultions to an entire process I don&#x27;t approve of, but maybe the fuzzy tools need old style deterministic tools the same way and for the same reasons we do.<p>So instead of an LLM trying to answer a math or reason question by finding a statistical match with other similar groups of words it found on 4chan and the all in podcast and a terrible recipe for soup written by a terrible cook, it can use a calculator when it needs a calculator answer.
      • stevula0 minutes ago
        I think that is how the smarter agents do things? Just like Claude&#x2F;ChatGPT sometimes does a web search they can do other tool calls instead of just making a statistical guess. Of course it doesn’t always make the bright choice between those options though…
      • colechristensen1 hour ago
        No, they just need to be trained to have adversarial self review &quot;thinking&quot; processes.<p>You ask an LLM &quot;What&#x27;s wrong with your answer?&quot; and you get pretty good results.
        • binary00101 hour ago
          Or you get the original output result was perfect and the adversarial &quot;rethinking&quot; switches to an incorrect result.
    • themafia1 hour ago
      &gt; we&#x27;re not actually on the right track to achieve real intelligence.<p>Real intelligence means you have to say &quot;I don&#x27;t know&quot; when you don&#x27;t know, or ask for help, or even just saying you refuse to help with the subtext being you don&#x27;t want to appear stupid.<p>The models could ostensibly do this when it has low confidence in it&#x27;s own results but they don&#x27;t. What I don&#x27;t know if it&#x27;s because it would be very computationally difficult or it would harm the reputation of the companies charging a good sum to use them.
      • wagwang26 minutes ago
        You can just tell the agent to do exactly that
        • alterom1 minute ago
          <i>&gt;You can just tell the agent to do exactly that</i><p>You can.<p>It just <i>won&#x27;t do it</i>.
      • cmrdporcupine54 minutes ago
        That&#x27;s just not how they work, really. They don&#x27;t know what they don&#x27;t know and their process requires an output.<p>I think they&#x27;re getting better at it, but it&#x27;s likely just the number of parameters getting bigger and bigger in the SOTA models more than anything.
        • adastra2217 minutes ago
          They do know what they don&#x27;t know. There&#x27;s a probability distribution for outputs that they are sampling from. That just isn&#x27;t being used for that purpose.
          • Isamu1 minute ago
            Oh, you mean somewhere it is tracking the statistical likelihood of the output. Yeah I buy that, although I think it just tends towards the most likely output given the context that it is dragging along. I mean it wouldn’t deliberately choose something really statistically unlikely, that’s like a non sequitur.
      • bluefirebrand1 hour ago
        My theory is because the people building the models and in charge of directing where they go love the sycophantic yes-man behavior the models display<p>They don&#x27;t like hearing &quot;I don&#x27;t know&quot;
      • colechristensen1 hour ago
        You can TELL the models to do this and they&#x27;ll follow your prompt.<p>&quot;Give me your answer and rate each part of it for certainty by percentage&quot; or similar.
        • mylifeandtimes50 minutes ago
          could you please tell me how it generates that certainty score?
          • adastra2216 minutes ago
            Vibes.
          • colechristensen32 minutes ago
            The whole thing is a statistical model, that&#x27;s just what it is. No, I cannot in a reasonable way dissect how an LLM works to a satisfactory level to a skeptic.
            • skydhash24 minutes ago
              It&#x27;s a statistical model for words and sentences, not knowledge. What does the LLM knows about having a pebble in your shoes, or drinking a nice cup of coffee?
  • dmix19 minutes ago
    &gt; They specifically address the AI Scribe program, the Ontario Ministry of Health initiated for physicians, nurse practitioners, and other healthcare professionals across the broader health sector.<p>makes me wonder what quality software the ministry would push (probably mostly qualifications like SOC).<p>This is apparently this list of approved vendors<p><a href="https:&#x2F;&#x2F;www.supplyontario.ca&#x2F;vor&#x2F;software&#x2F;tender-20123-artificial-intelligent-solutions-ai-scribe&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.supplyontario.ca&#x2F;vor&#x2F;software&#x2F;tender-20123-artif...</a>
  • Hobadee2 hours ago
    The AI note taker we use at work records the meeting as well, and each note it takes about the meeting has a timestamp link that takes you directly there in the recording so you can check it yourself. While I&#x27;m sure a solution like this is more complicated in a HIPPAA environment, something like this is critical for things as important as healthcare.
    • TonyAlicea101 hour ago
      When designing AI-based user experiences I refer to this as provenance. It’s a vital aspect of trust, reliability, compliance and more. If a software system includes LLM output like this but doesn’t surface the provenance of its output for human evaluation and verification then it’s at best poor user experience, and at worst a dangerous one.
      • autoexec57 minutes ago
        At the same time, do you really want every conversation you have with your doctor recorded, handed over to third party companies, and stored forever with your medical file? Plus what doctor has time to sit down and re-listen to your visit to check to make sure the AI didn&#x27;t screw up at some point in the future anyway? If your doctor isn&#x27;t going to be verifying the accuracy from those recordings who would? Overseas contractors? At what point does it become a larger waste of time and money to babysit an incompetent AI than just not using one in the first place?<p>There are some good uses for AI, but I&#x27;m not convinced that this (or many other cases where accuracy matters) is one of them.
  • ceejayoz2 hours ago
    &gt; 60% of evaluated AI Scribe systems mixed up prescribed drugs in patient notes, auditors say<p>Not mentioned, as far as I can see: the comparative human mistake rate.<p>Having seen a <i>lot</i> of medical records, 60% sounds about normal lol.
    • autoexec49 minutes ago
      Even if you had the same 60% error rate with humans the types of errors would be vastly different. Humans might make typos, or forget to include something, or even occasionally misremember some minor detail, but that&#x27;s very different from BS AI just hallucinates out of nowhere. AI makes the kinds of mistakes no human ever would which means they can be extremely confusing and easy to catch or they can be something no human would even think to question or be looking out for because it makes no sense why AI would randomly (and confidently) say something so wrong.
    • thepotatodude1 hour ago
      60% is insanely high and absolutely not the performance of human mistake rate. What charts are you reading?
    • vor_5 minutes ago
      60% is a normal human mistake rate? You can&#x27;t be serious.
    • Arodex2 hours ago
      But who is responsible is different.<p>(And if you already see 60% error rates in standard, pre-AI note taking, how does that not translate into many deaths and injury? At least one country&#x27;s health system in the world should have caught that)
      • tredre31 hour ago
        &gt; And if you already see 60% error rates in standard, pre-AI note taking, how does that not translate into many deaths and injury?<p>Presumably most doctor&#x27;s visits are a one-problem-one-solution-one-doctor type of thing. Done deal, notes are never read again. So that alone would explain why high rates of errors doesn&#x27;t result in injuries or death very often.<p>Any injury or death caused by poor notes would have to occur when mistakes are done if you&#x27;re followed for a serious chronic condition, or if you&#x27;re handled by a team where effective communication is required.
      • ceejayoz2 hours ago
        &gt; how does that not translate into many deaths and injury?<p>Because most of it is just written down and never looked at again until there’s a lawsuit or something.
      • cyanydeez2 hours ago
        Yeah, the problem is the health system has no sacrificial goat if the AI note taker provides the wrong detail. The last thing we want is CTO being responsible!
        • bluefirebrand2 hours ago
          I&#x27;m not convinced the CTO would be held accountable either.<p>I do wonder if people would be pushing AI so hard if their organizations were planning to hold them accountable for mistakes the AI made<p>I bet if that were the case we&#x27;d see a lot slower rollout of AI systems
    • jmward011 hour ago
      This is not a popular view &#x27;AI sucks at X but so do humans&#x27; but I think it is valid and we should take wins where we can, especially in healthcare. It is pretty clear that initial accuracy issues will become less and less of a problem as these technologies mature. This focus on accuracy now as a &#x27;see it&#x27;s bad&#x27; talking point though misses the real danger. Medical note takers have an exceptionally high chance of being hijacked for money and that is an issue we need to bring attention to now. They provide a real-time feed into a trillion dollar industry. Just roll that around in your head for a second. Insurance companies are going to want to tap that feed in real time so they can squeeze more money out. Drug makers are going to want to tap into that feed so they can abuse the data. Hospitals will want to tap into that feed to wring more out of doctors and boost the number of billable codes for each encounter. Very few entities are looking to tap into that feed to, you guessed it, help the patient. I am for these systems (and I have been involved in building them in the past) but the feeding frenzy of business interest that will obviously get involved with them is the thing we should be yelling and screaming about, not short-term accuracy issues.
      • NateEag5 minutes ago
        &gt; It is pretty clear that initial accuracy issues will become less and less of a problem as these technologies mature.<p>What do you base this on?<p>As someone who can both see the amazing things genAI can do, and who sees how utterly flawed most genAI output is, it&#x27;s not obvious to me.<p>I&#x27;m working with Claude every day, Opus 4.7, and reviewing a steady stream of PRs from coworkers who are all-in, not just using due to corporate mandates like me, and I find an unending stream of stupidity and incomprehension from these bots that just astonishes me.<p>Claude recently output this to me:<p>&quot;I&#x27;ve made those changes in three files:<p>- File 1<p>- File 2&quot;<p>That is a vintage hallucination that could&#x27;ve come right out of GPT 2.0.
      • mcphage34 minutes ago
        &gt; It is pretty clear that initial accuracy issues will become less and less of a problem as these technologies mature.<p>Does it?
  • nothinkjustai1 hour ago
    People will eventually figure out LLMs have no capacity for intent and are fundamentally unreliable for tasks such as summarization, note taking etc.
  • jqpabc12345 minutes ago
    And once again, we have an example of how AI is a liability issue waiting to happen.
  • samcooper1 hour ago
    [flagged]