Artificial intelligence is worse than humans in every way at summarising documents and might actually create additional work for people, a government trial of the technology has found.

Amazon conducted the test earlier this year for Australia’s corporate regulator the Securities and Investments Commission (ASIC) using submissions made to an inquiry. The outcome of the trial was revealed in an answer to a questions on notice at the Senate select committee on adopting artificial intelligence.

The test involved testing generative AI models before selecting one to ingest five submissions from a parliamentary inquiry into audit and consultancy firms. The most promising model, Meta’s open source model Llama2-70B, was prompted to summarise the submissions with a focus on ASIC mentions, recommendations, references to more regulation, and to include the page references and context.

Ten ASIC staff, of varying levels of seniority, were also given the same task with similar prompts. Then, a group of reviewers blindly assessed the summaries produced by both humans and AI for coherency, length, ASIC references, regulation references and for identifying recommendations. They were unaware that this exercise involved AI at all.

These reviewers overwhelmingly found that the human summaries beat out their AI competitors on every criteria and on every submission, scoring an 81% on an internal rubric compared with the machine’s 47%.

  • maegul (he/they)
    link
    fedilink
    English
    arrow-up
    23
    arrow-down
    1
    ·
    1 month ago

    Not a stock market person or anything at all … but NVIDIA’s stock has been oscillating since July and has been falling for about a 2 weeks (see Yahoo finance).

    What are the chances that this is the investors getting cold feet about the AI hype? There were open reports from some major banks/investors about a month or so ago raising questions about the business models (right?). I’ve seen a business/analysis report on AI, despite trying to trumpet it, actually contain data on growing uncertainties about its capability from those actually trying to implement, deploy and us it.

    I’d wager that the situation right now is full a lot of tension with plenty of conflicting opinions from different groups of people, almost none of which actually knowing much about generative-AI/LLMs and all having different and competing stakes and interests.

    • Voroxpete@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      18
      arrow-down
      2
      ·
      1 month ago

      “What are the chances…”

      Approximately 100%.

      That doesn’t mean that the slide will absolutely continue. There may be some fresh injection of hype that will push investor confidence back up, but right now the wind is definitely going out of the sails.

      The core issue, as the Goldman - Sachs report notes, is that AI is currently being valued as a trillion dollar industry, but it has not remotely demonstrated the ability to solve a trillion dollar problem.

      No one selling AI tools is able to demonstrate with confidence that they can be made reliable enough, or cheap enough, to truly replace the human element, and without that they will only ever be fun curiosities.

      And that “cheap enough” part is critical. It is not only that GenAI is deeply unreliable, but also that it costs a truly staggering amount of money to operate (OpenAI are burning something like $10 billion a year). What’s the point in replacing an employee you pay $10 an hour to handle customer service issues with a bot that costs $5 for every reply it generates?

      • kautau@lemmy.world
        link
        fedilink
        English
        arrow-up
        8
        ·
        1 month ago

        Yeah we are on the precipice of a massive bubble about to burst because, like the dot com bubble magic promises are being made by and to people who don’t understand the tech as if it is some magic that will net incredible profits just by pursuing it. LLMs have great applications in specific things, but they are being thrown in every direction to see where they will stick and the magic payoff will come

        • Voroxpete@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          4
          arrow-down
          1
          ·
          1 month ago

          The problem is that even the specific things they’re good at, they don’t do well enough to justify spending actual money on. And when I say “actual money”, I’m not talking about the hilariously discounted prices AI companies are offering in an effort to capture an audience.

          A bot that can do a job reasonably well, but still needs a human to check their work is, from an employment perspective, still an employee, just now with some very expensive helper software. And because of the inherent unreliability of LLMs, a problem that many top figures in the industry are finally admitting may never be solved, they will always need a human to check their work. And that human has to be competent enough to do the job without the AI, in order to figure out where and how it went wrong.

          GenAI was supposed to put us all out of work, and maybe one day it will, but the current state of the technology isn’t remotely close to being good enough to do that. It turns out that while bots can very effectively look and sound like humans, they’re not remotely capable of thinking like humans, and that actually matters when your chatbot starts promising customers discounts that don’t actually exist, to name one real example. What was treated as being the last ten percent is actually looking more and more like ninety-nine percent of the work in terms of creating something that can effectively replace a human being.

          (As an aside, I can’t help but feel that a big part of this epic faceplant arises from Silicon Valley fully ingesting the bullshit notion of “unskilled labour”. Turns out working the drive thru at McDonald’s is a more complicated job than people think, including McDonald’s themselves. We’ve so undervalued the skills of vast swathes of our population that we were easily deluded into thinking they could all be replaced by simple machines. While some of those tasks certainly can, and will, be automated, there are some human elements - especially in conflict resolution - that are really hard to replace)

      • maegul (he/they)
        link
        fedilink
        English
        arrow-up
        4
        ·
        1 month ago

        Yea, the “cheaper than droids” line in Andor feels strangely prescient ATM.

    • Optional@lemmy.world
      link
      fedilink
      English
      arrow-up
      6
      ·
      1 month ago

      What are the chances that this is the investors getting cold feet about the AI hype?

      Investors have proven over and over they’re credulous idiots who understand sweet fuck-all about technology and will throw money at whatever’s in their face. Creepy Sam and the Microshits will trot out some more useless garbage and prize a few more billion out of the market in just a little while.

    • atrielienz@lemmy.world
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      2
      ·
      edit-2
      1 month ago

      NVIDIA has been having a lot of problems with their 13th/14th gen CPU’s degrading. They are also embroiled in an anti-trust investigation. That coupled with the “growing pains of generative AI” has caused them a lot of problems where 2 months ago they were one of the world’s most valuable companies.

      Some of it is likely the die-off of the AI hype but their problems are farther reaching than the sudden AI boom.