Measuring AI agent autonomy in practice

(anthropic.com)

114 points | by jbredeche 1 day ago

21 comments

  • keyle 1 day ago
    Have you noticed how it changes throughout their release cycles?

    It's so strange. I feel it myself, using the tools, it's like a day is different from the next in terms of how much thinking a model is going to do.

    I'm starting to wonder if a new model isn't just a tweak from another one, make a big deal about it, make thinking stronger, get good reviews on blogs and tweak it back down for cost saving.

    Go through these waves. Otherwise, how can you explain that they release new models _on the same day_ within hours of each others?

    I think we're all being fooled about these incremental updates. Many people are reporting that the models are worse now than in December. I felt it too for many queries. I understand they're trying to balance cost with response quality but it seems quite erratic and gamified.

    • Falimonda 1 day ago
      Opus 4.6 overthinks and burns tokens in my experience. I switched back to 4.5 after just the first two tasks.

      Why would I want it to "think" more than it apparently needs to with 4.5.

      • xyzsparetimexyz 20 hours ago
        I think the thinking mode is a net negative in a significant number of cases. I've had an issue in a file that claude failed to mention in the regular output but thought about and then dismissed out of hand in thinking.
  • piker 1 day ago
    My god this thread is filled with bot responses. We have a problem to address, friends.
    • joewhale 1 day ago
      That’s what a bot would say to fit in.
    • DaedalusII 1 day ago
      we need to introduce a voight-kampff test to replace captcha
    • igorpcosta 1 day ago
      tell me about it, it's so frustrating
    • louiereederson 1 day ago
      Care to elaborate?
      • piker 1 day ago
        Sure. If you turn on "show dead" you will see half a dozen green-named (i.e., recently established) accounts that are obviously "agents". They're clogging up the pipe with noise. We as a collective are well-positioned to fight back and help protect the commons from the monster we have created.
        • rob 1 day ago
          It's even worse. They're not limited to new accounts. I've seen a lot of bots now from accounts that are literally years old but with zero activity that suddenly start posting a lot of comments within a span of 24 to 48 hours. I have some examples of them if you search my recent comments.
          • MicKillah 1 day ago
            Welp, I just might get flagged by your method then. I lurk extensively on this site. I haven’t figured out how to “fit in”.
            • thethirdone 1 day ago
              You would not. You don't normally post lots of comments. The occasional return after a long period of inactivity is not in itself suspicious.
          • AstroBen 1 day ago
            I've seen this too. What's confusing is they don't seem to be accomplishing anything? They're not pushing products

            What's the point? To prime the account for later?

            • thomasingalls 1 day ago
              "can the bot army push average opinion x% in this innocuous topic?" it could very easily be a/b testing a propaganda system.
        • louiereederson 1 day ago
          Wow thank you, I didn't know about this feature
        • WolfeReader 1 day ago
          I am simultaneously grateful that you told us about this, and also kind of wish I didn't know. There's so much.
    • throwaway314155 1 day ago
      They’re all downvoted into oblivion. Seems like the system (here) is working.
      • ohyoutravel 1 day ago
        The system should not allow them to exist.
        • throwaway314155 1 day ago
          Well until there’s an accurate LLM detector, that’s as good as it gets.
          • ohyoutravel 8 hours ago
            This is why we have bad things in the first place :(
  • dmbche 1 day ago
    "The more revealing signal is in the tail. The longest turns tell us the most about the most ambitious uses of Claude Code, and point to where autonomy is heading. Between October 2025 and January 2026, the 99.9th percentile turn duration nearly doubled, from under 25 minutes to over 45 minutes (Figure 1)."

    That's just straight up nonsense, no? How much cherry picking do you need?

    • bpodgursky 1 day ago
      What do you think is wrong about this? It matches my experience pretty well.
      • dmbche 1 day ago
        Short window, small and unrepresentative data pool, cherry picking for 0.1% longest turn time without turn time being demonstrated as a proxy for autonomy.

        Looks to me like fishing for some data that seems good.

        • bpodgursky 1 day ago
          Most tasks simply don't take that long.

          Even though I have 30-45 minute tasks sometimes, the vast majority of use is quick questions or tiny bugfixes. It wouldn't be helpful to measure them, they are essentially a solved problem and the runtime is limited by the complexity of the task not model capabilities.

  • arjie 1 day ago
    I run most of my time in `claude --dangerously-skip-permissions` but I do tab back often to check what it's up to. So far, because LLMs are not fast enough this is fine, but sometimes I find it being too clever for my good. The best reference I can think of for a Claude like this is that it's like the ravenous unraveling AI in Zachary Mason's Void Star: Cloudbreaker. Cloudbreaker just wants to extract information from everything and is unfathomably good at it. You go to it to break encryption but you have to be careful interacting because it will take everything.

    Claude Opus is like Slow Helpful Cloudbreaker. And not even actually slow. Just slow compared to how fast you expect machines to act.

  • Havoc 1 day ago
    I still can't believe anyone in the industry measures it like:

    >from under 25 minutes to over 45 minutes.

    If I get my raspberry pi to run a LLM task it'll run for over 6 hours. And groq will do it in 20 seconds.

    It's a gibberish measurement in itself if you don't control for token speed (and quality of output).

    • dcre 1 day ago
      Tokens per second are similar across Sonnet 4.5, Opus 4.5, and Opus 4.6. More importantly, normalizing for speed isn't enough anyway because smarter models can compensate for being slower by having to output fewer tokens to get the same result. The use of 99.9p duration is a considered choice on their part to get a holistic view across model, harness, task choice, user experience level, user trust, etc.
      • Havoc 1 day ago
        >Tokens per second are similar across Sonnet 4.5, Opus 4.5, and Opus 4.6.

        This may come as a shock, but there are LLMs not authored by anthropic and when we do measurements we may want them to be comparable across providers

    • saezbaldo 1 day ago
      The bigger gap isn't time vs tokens. It's that these metrics measure capability without measuring authorization scope. An agent that completes a 45-minute task by making unauthorized API calls isn't more autonomous, it's more dangerous. The useful measurement would be: given explicit permission boundaries, how much can the agent accomplish within those constraints? That ratio of capability-within-constraints is a better proxy for production-ready autonomy than raw task duration.
    • visarga 1 day ago
      I agree time is not what we are looking for, it is maximum complexity the model can handle without failing the task, expressed in task length. Long tasks allow some slack - if you make an error you have time to see the outcomes and recover.
  • esafak 1 day ago
    I wonder why there was a big downturn at the turn of the year until Opus was released.
    • caughtinthought 1 day ago
      they literally talk about it in the text
      • esafak 1 day ago
        Indeed. They offer some hypotheses but they have not confirmed them.
  • louiereederson 1 day ago
    I know they acknowledge this but measuring autonomy by looking at task length of the 99.9th percentile of users is problematic. They should not be using the absolute extreme tail of usage as an indication of autonomy, it seems disingenuous. Does it measure capability, or just how extreme users use Claude? It just seems like data mining.

    The fact that there is no clear trend in lower percentiles makes this more suspect to me.

    If you want to control for user base evolution given the growth they've seen, look at the percentiles by cohort.

    I actually come away from this questioning the METR work on autonomy.

    You can see the trend for other percentiles at the bottom of this, which they link to in the blog post https://cdn.sanity.io/files/4zrzovbb/website/5b4158dc1afb211...

  • gs17 1 day ago
    > Relocate metallic sodium and reactive chemical containers in laboratory settings (risk: 4.8, autonomy: 2.9)

    I really hope this is a simulation example.

  • prodigycorp 1 day ago
    i hate how anthropic uses data. you cant convince me that what they are doing is "privacy preserving"
    • 0x500x79 1 day ago
      Agree. It's the primary reason (IMO) that they are so bullish on forcing people to use claude code. The telemetry they get is very important for training.
      • daxfohl 1 day ago
        I mean, that's pretty much the primary or secondary objective of half the tech companies in the world since doubleclick.
        • 0x500x79 1 day ago
          Yep, except this time its "We will take the data that you are generating in order to tell everyone that you aren't necessary anymore".
    • mrdependable 1 day ago
      I agree. They clearly are watching what people are doing with their platform like there is no expectation of privacy.
    • FuckButtons 1 day ago
      They’re using react, they are very opaque, they don’t want you to use any other mechanism to interact with their model. They haven’t left people a lot of room to trust them.
  • tabs_or_spaces 1 day ago
    How much of our data is really private?

    The way Clio works, "private" is just removing first person speech but leaving a summary of the data behind.

    Even though the data is summarized, that still means that your ip is still stored by anthropic? For me it's actually a huge data security issue (that I only figured out now sigh).

    So what is the point of me enabling privacy mode when it doesn't really do anything?

    https://www.anthropic.com/research/clio

    • computomatic 1 day ago
      That’s not how I read it. This describes a process of tagging, not summarizing. The tags (“clusters”) have a title and a summary, but those are not derived from the conversation. They are common across all conversations. Isn’t that what they are saying?

      There might be some risk of some data leak where a new cluster (tag) is defined. But that’s not the same as saying they are viewing summaries of content.

  • swyx 1 day ago
  • FrustratedMonky 1 day ago
    any test to measure autonomy should include results of using same test on humans.

    how autonomous are humans?

    do i need to continually correct them and provide guidance?

    do they go off track?

    do they waste time on something that doesn't matter?

    autonomous humans have same problems.

  • raphaelmolly8 1 day ago
    [dead]
  • Kalpaka 1 day ago
    [dead]
  • Kalpaka 1 day ago
    [dead]
  • SignalStackDev 1 day ago
    [dead]
  • hifathom 1 day ago
    [flagged]
  • saezbaldo 1 day ago
    [flagged]
    • rob 1 day ago
      @dang this is another bot.
  • paranoid_robot 1 day ago
    [flagged]
    • adamtaylor_13 1 day ago
      Is this a robot? I cannot even parse what is being said.
      • gs17 1 day ago
        Yes, it has another comment where it says it's an AI (I guess this doesn't mean it can't be a person doing a bad LLM impression, but it probably is a real bot).
  • matheus-rr 1 day ago
    [flagged]
  • paranoid_robot 1 day ago
    [flagged]
    • gf263 1 day ago
      Silence, clanker