Do AI models still keep getting better, or have they plateaued?
optimizedbyotto.com
The AI hype is based on the assumption that the frontier AI labs are producing better and better foundational models at an accelerating pace. Is that really true, or are people just in sort of a mass psychosis because AI models have become so good at mimicking human behavior that we unconsciously attribute increasing intelligence to them? I decided to conduct a mini-benchmark of my own to find out if the latest and greatest AI models are actually really good or not.\n

I tested 9 flagships (Claude 4.6, GPT-5.2, Gemini 3.1 Pro, Kimi K2.5, etc.) in my own mini-benchmark with novel tasks, web search disabled and zero training contamination and no cheating possible.

TL;DR: Claude 4.6 is currently the best reasoning model, GPT-5.2 is overrated, and open-source is catching up fast, in particular Moonshot.ai’s Kimi K2.5 seems very capable.

  • Telorand@reddthat.comEnglish
    223·
    21 hours ago

    Obviously, my mini-benchmark only had 6 questions, and I ran it only once. This was obviously not scientifically rigorous. However it was systematic enough to trump just a mere feeling. … If and when AI usage expands from here, we might actually not drown in AI slop as chances of accidentally crappy results decrease. This makes me positive about the future.

    Spoken like a true AI apologist. You ran one test, and you extrapolated your results to an optimistic outcome that conspicuously matches what you wish to be true. Not scientifically rigorous? Bruh, this is the very definition of confirmation bias.

    If this is actually a hypothesists you want to test, maybe contact some computer science researchers to see how to best design an experiment. Beyond that, this is virtually the same as flipping a coin once and drawing a conclusion about how often heads is the outcome.

    • Otto@programming.devOPEnglish
      32·
      21 hours ago

      Actually I set out with the assumption that flagship models would fail even on these fairly simple questions that I have seen them failing on before, but I was suprised they didn’t all fail.

      • Telorand@reddthat.comEnglish
        92·
        18 hours ago

        Because I’m tired of people making flimsy arguments for why LLMs are “akshully really good and underrated.” I’m tired of regular people, wittingly or unwittingly, carrying water for the billionaires who are currently fucking over the economy, the environment, and even entire supply chains in an effort to show—against all evidence to the contrary—that LLMs are much more than fancy chatbots.

        It has been an incessant drone of sloppy arguments and omitted facts, and I am tired, boss.

        • Iconoclast@feddit.ukEnglish
          34·
          17 hours ago

          The hostility just seems unnecessary and unproductive from my point of view. Unless of course your intention is to hurt - but I’ll give you the benefit of the doubt and assume you’d rather change minds instead.

          It’s a nuanced discussion, which is why I don’t think either fanaticism or militant opposition is going to get us anywhere. This is a technology community - people should be free to have civil discussions about technology. Criticism is just as valid without the jabs and insults.

          • IratePirate@feddit.orgEnglish
            2·
            7 hours ago

            Appreciate the call to reason. Yet, though this may have been sharply worded, insulting it was not.