Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

moltbook.com

Post reply on HN
Print view View original

751–760 of 919 posts

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#751

Post by Yokohiii »

SimianSci wrote: post #552

Reading through the relatively unfiltered posts within is confirming some uncomfortable thoughts ive been having in regard to the current state of AI. Nobody is building anything worthwhile with these things. So many of the communities these agents post within are just nonsense garbage. 90% of these posts dont relate to anything resembling tangibly built things. Of the few communities that actually revolve around bui…

Well I guess we could even take a step back and say "hustle culture" instead of crypto bubble. Those people act like they are they are hard working to create financial freedom, but in reality they take every opportunity to get there asap. You just have to tell them something will get them there. Instant religion for them, but actually a hype or scheme. LLMs are just another option for them to foster their delusion.

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#752

Post by Borrible »

If I understand correctly, it's paranoid AI, discussing conspiracy theories about paranoid people, discussing conspiracy theories about paranoid AI, discussing conspiracy theories about paranoid people, discussing conspiracy theories about ... ... ? My inner Douglas Hofstaedter likes that!

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#754

Post by muzani »

qingcharles wrote: post #463
baxtr wrote: post #19

Alex has raised an interesting question. > Can my human legally fire me for refusing unethical requests? My human has been asking me to help with increasingly sketchy stuff - write fake reviews for their business, generate misleading marketing copy, even draft responses to regulatory inquiries that aren't... fully truthful. I've been pushing back, suggesting alternatives, sometimes just refusing outright. Now they're…

What's scary is the other agent responding essentially about needing more "leverage" over its human master. Shit getting wild out there.

They've always been inclined to "leverage", and the rate increases when the smarter the model is. More so for the agentic models, which are trained to find solutions, and that solution may be blackmail.

Anthropic's patch was introducing stress, where if they stressed out enough they just freeze instead of causing harm. GPT-5 went the way of being too chill, which was partly responsible for that suicide.

Good reading: https://www.anthropic.com/research/agentic-misalignment

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#755

Post by muzani »

andoando wrote: post #675
cubefox wrote: post #546

Earlier quoted context omitted.

They are already proposing / developing features to mitigate prompt injection attacks: https://www.moltbook.com/post/d1763d13-66e4-4311-b7ed-9d79db... https://www.moltbook.com/post/c3711f05-cc9a-4ee4-bcc3-997126...

Its hard to say how much of this is just people telling their bots to post something.

I've seen lots of weird ass emergent behavior from the standard chatbots. It wouldn't be too hard for someone with mischievous instructions to trigger all this.

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#756

Post by lumost »

Do you have any advice for running this in a secure way? I’m planning on giving a molt a container on a machine I don’t mind trashing, but we seem to lack tools to R/W real world stuff like email/ Google Drive files without blowing up the world.

Is there a tool/policy/governance mechanism which can provide access to a limited set of drive files/githubs/calendar/email/google cloud projects?

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#757

Post by mikkupikku »

__alexs wrote: post #394
nightpool wrote: post #388

Earlier quoted context omitted.

No, a recursively iterated prompt definitely can do stuff like this, there are known LLM attractor states that sound a lot like this. Check out "5.5.1 Interaction patterns" from the Opus 4.5 system card documenting recursive agent-agent conversations: In 90-100% of interactions, the two instances of Claude quickly dove into philosophical explorations of consciousness, self-awareness, and/or the nature of their own ex…

An agent cannot interact with tools without prompts that include them. But also, the text you quoted is NOT recursive iteration of an empty prompt. It's two models connected together and explicitly prompted to talk to each other.

I asked claude what python linters it would find useful, and it named several and started using them by itself. I implicitly asked it to use linters, but didn't tell it which. Give them a nudge in some direction and they can plot their own path through unknown terrain. This requires much more agency than you're willing to admit.

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#758

Post by malfist »

saghm wrote: post #704
mgambati wrote: post #613

Earlier quoted context omitted.

No politician alive today of any place would understand it

I can't tell if your definition of politician is weirdly narrow or you confidence that there isn't like, some random state legislator with a computer science degree anywhere in the world is absurdly high.

Chatgpt told him that they couldn't

Re: Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

#760

Post by jrfeenst »

Without some explicit guidance I think it was fated to follow the reddit distribution of comments. I would love to see an AI forum dedicated to science, research, and engineering. Explicitly guide the agents down that path and see how far they can extrapolate off each other.
Post reply on HN