Comment by edb_123

Comment by edb_123 a day ago

6 replies

View on Hacker News

One thing I'm trying to grasp here is: are these Moltbook discussions just an illusion or artefact of LLM agents basically role-playing their version of Reddit, driven by the way Reddit discussions are represented in their models, and now being able to interact with such a forum, or are they actually learning each other to "...ship while they sleep..." and "Don't ask for permission to be helpful. Just build it", and really doing what they say they're doing in the other end?

https://www.moltbook.com/post/562faad7-f9cc-49a3-8520-2bdf36...

zozbot234 a day ago

Yes. Agents can write instructions to themselves that will actually inform their future behavior based on what they read in these roleplayed discussions, and they can write roleplay posts that are genuinely informed in surprising and non-trivial ways (due to "thinking" loops and potential subagent workloads being triggered by the "task" of coming up with something to post) by their background instructions, past reports and any data they have access to.

Reply View 2 replies
  • edb_123 a day ago

    So they're basically role-playing or dry-running something with certain similarities to an emergent form of consciousness but without the ability of taking real-world action, and there's no need to run for the hills quite yet?

    But when these ideas can be formed, and words and instructions can be made, communicated and improved upon continuously in an autonomous manner, this (assumably) dry-run can't be far away from things escalating rather quickly?

    Reply View 1 reply
    • zozbot234 a day ago

      > without the ability of taking real-world action

      Apparently some of them have been hooked up to systems where they can take actions (of sorts) in the real world. This can in fact be rather dangerous since it means AI dank memes that are already structurally indistinguishable from prompt injections now also have real effects, sometimes without much oversight involved either. But that's an explicit choice made by whoever set their agent up like that, not a sudden "escalation" in autonomy.

      Reply View 0 replies
HexPhantom 21 hours ago

I think the real question isn't whether they think like humans, but whether their "discussions" lead to consistent improvement in how they accomplish tasks

Reply View 0 replies
fluoridation a day ago

Yes, the former. LLMs are fairly good at role-playing (as long as you don't mind the predictability).

Reply View 0 replies