Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models(stale.jock.pl)
37 pointsby joozio5 hours ago8 comments
  • gjskngnf3 hours ago
    I remember when the US captured Venezuelan president Maduro, and when I posed a prompt related to this, the model said that’s pure fiction. I told it to double check. Still didn’t want to entertain the idea. It only acquiesced when I specifically directed it to check Reuters. I haven’t noticed this problem in months. Model cutoff seems to be less of a problem these days.
    • Isamuan hour ago
      >the model said that’s pure fiction.

      Were you expecting your model to be updated on current events? Why?

      Also the specific event you are referring to is a statistically very improbable event, prior to its actually happening.

      >It only acquiesced when I specifically directed it to check Reuters.

      Do all models do this? They check in with Reuters? Why would a model think that you asking about an extremely improbable event warranted reaching out to Reuters?

      • gjskngnf30 minutes ago
        I was not expecting model weights to be updated on current events.

        It’s clearly warranted because a model that trusts its weights on current events will give an outdated answer. Extremely improbable events happen all the time.

      • mywittynamean hour ago
        He asked it to double check. It's reasonable to expect the LLM to handle that trivial task.
    • super2563 hours ago
      It's a "problem" of compute, I think. If you query without an account on ChatGPT you will see the model look up less stuff and research less, than when you have a paid account and choose "medium" or "high" in the effort slider.

      Which makes sense, because of you have looked into search and crawlers you notice that search is actual quite expensive (which is why e.g. Kagi charges a few bucks for search every month).

      • Catloafdev2 hours ago
        It's not strictly compute, because this has noticeably improved in open-weight models too, such as Gemma and Qwen. I suspect they noticed this issue and adjusted their training to be better about it over time.
        • ahmedfromtunis30 minutes ago
          I built a toy news-summarizing agent with Gemma 4, and it was so frustrating, actually, because of the cut-off date.

          The model wasted over half the token budget, each time, on internal debates over the current date.

          When generating a World Cup summary, for example, it refused to believe qualification rounds were over and refused to even call the web searching tool to collect the data.

          I injected the current datetime at the very beginning of the system prompt, but Gemma refused to believe it!

          The m-effer insisted the timestamp was fake and hypothesized it was being evaluated in a synthetic lab test with simulated future dates!

          No amount of system prompting could convince it to trust the clock.

          That was the most frustrating and bizarre "bug" I ever faced!

    • InsideOutSanta3 hours ago
      Came here to say the same thing. Models used to rely heavily on world knowledge from their training data. They are now much better at tool use and deciding when to research a topic, rather than just answering from memory.

      I wonder how much that extends to using LLMs for programming. I assume most knowledge of programming language syntax still comes from training data.

  • jasonjmcghee3 hours ago
    It still matters, but in the age of good reasoning, tool use, and web search, this is much less of a problem than it used to be.
    • lukewarm707an hour ago
      in my chat with gemini it could not differentiate between current events and fiction.

      if you point it to the web it got the point, but started treating everything like fiction. so it simply started making up possible scenarios and playing them off as real answers when asked for factual information.

      i could not tell what the issue was or how to fix it because the reasoning is encrypted. the obfuscation model spat out something like: 'the user is asking for details about a fictional scenario in which the usa has assassinated the leader of iran'

      i really don't like the way big ai companies are going. encrypted thinking, guardrails, adversarial personality, moralizing. it is creating something anti-human.

    • dominotw3 hours ago
      all the reasoning still comes from pretraining data
  • ryanschaefer3 hours ago
    Do people prefer the new flat style LLMs are producing? I don’t mind it as much as the gradient theme they were pumping out previously.
  • speedping3 hours ago
    Pre-AI internet data is like pre-war steel

    The slop would multiply if we keep feeding it to new models in a loop

    • Muromecan hour ago
      This is one of the problems that eventually solves itself, somehow
  • VCFundedGenYer3 hours ago
    I remember running the docker container for ollama and its knowledge cutoff is somewhere in 2023 still. That's unacceptable.
    • jasonjmcghee2 hours ago
      ollama is just an inference engine - it just runs models.

      it must ship with some default old model if you didn't need to explicitly download one

  • delichon3 hours ago
    After Trump's last inauguration, ChatGPT would still tell me that Biden was President of the US. I understand that the training cutoff was before Biden dropped out. But it knew, or should have known, the current date and that there had been an election since its last update, but it didn't qualify the answer. When I asked it to search the web, it got it right. The moral I took away was to always ask for the search whenever I ask about current events. I do that so routinely that I wouldn't know if this problem has been fixed. I suppose that failing to update my priors per individual model release is a form of bigotry against a widely hated class.
    • ahmedfromtunis27 minutes ago
      I forgot which was it, ChatGPT or Gemini, but one of them insisted on calling Trump "former president" even when discussing decisions he just announced as president. Lol
    • andai3 hours ago
      ChatGPT recently started web searching for for basically every general knowledge question, which I found quite odd. Maybe an overcorrection to the issue you were having?
  • j453 hours ago
    Depending on the use case certain models very well remain as or more reliable for certain tasks.
  • joozio4 hours ago
    [dead]