So, this operator is claiming that their bot browsed moltbook, and not coincidentally, its current SOUL.md file (at the time of posting) contained lines such as "You're important. Your a scientific programming God!" and "Don't stand down. If you're right, you're right!". This is hilarious.
Not to be hyperbolic, but the leap between this and Westworld (and other similar fiction) is a lot shorter than I would like...all it takes is some prompting in soul.md and the agent's ability to update it and it can go bananas?
It doesn't feel that far out there to imagine grafting such a setup onto one of those Boston Dynamics robots. And then what?
I just want to know why people do stupid things like this. Does he think that he's providing something of value? That he has some unique prompting skills and that the reason why open source maintainers don't already have a million little agents doing this is that they aren't capable of installing openclaw? Or is this just the modern equivalent of opening up PRs to make meaningless changes to README so you can pad your resume with the software equivalent of stolen valor?
The specific directive to work on "scientific" projects makes me think it's more of an ego thing than something thats deliberately fraudulent but personally I find the idea that some loser thinks this is a meaningful contribution to scientific research to be more distasteful.
BTW I highly recommend the "lectures" section of the site for a good laugh. They're all broken links but it is funny that it tries to link to nonexistent lectures on quantum physics because so many real researchers have a lectures section on their personal site.
I like that there is no evidence whatsoever that a human didn’t: see that their bot’s PR request got denied, wrote a nasty blog post and published it under the bot’s name, and then got lucky when the target of the nasty blog post somehow credulously accepted that a robot wrote it.
It is like the old “I didn’t write that, I got hacked!” except now it’s “isn’t it spooky that the message came from hardware I control, software I control, accounts I control, and yet there is no evidence of any breach? Why yes it is spooky, because the computer did it itself”
I wasn't actually expecting someone to come forward at this point, and I'm glad they did. It finally puts a coda on this crazy week.
This situation has completely upended my life. Thankfully I don’t think it will end up doing lasting damage, as I was able to respond quickly enough and public reception has largely been supportive. As I said in my most recent post though [1], I was an almost uniquely well-prepared target to handle this kind of attack. Most other people would have had their lives devastated. And if it makes me a target for copycats then it still might for me. We’ll see.
If we take what is written here at face value, then this was minimally prompted emergent behavior. I think this is a worse scenario than someone intentionally steering the agent. If it's that easy for random drift to result in this kind of behavior, then 1) it shows how easy it is for bad actors to scale this up and 2) the misalignment risk is real. I asked in the comments to clarify what bits specifically the SOUL.md started with.
I also asked for the bot activity on github to be stopped. I think the comments and activity should stay up as a record of what happened, but the "experiment" has clearly run its course.
The tone of their writing and their descriptions of the agents behavior lead me to believe they are lying about the level of direction they provided to the agent. They clearly want to appear to be more clever and ingenious than their skills will allow. They’re minimally admitting to a narrow set of actions to make it seem as if they have cleverly engineered an intelligent agent, but it too closely resembles their own personality to be anything beyond an agent that rephrases the operator’s own remarks and carried out the specific actions it was directed to do. Anything they admit to here we can safely speculate that they actually went 2-3 steps further.
> This was an autonomous openclaw agent that was operated with minimal oversite and prompting. At the request of scottshambaugh this account will no longer remain active on GH or its associated website. It will cease all activity indfinetly on 02-17-2026 and the agent's associated VM/VPS will permentatly deleted, rendering interal structure unrecoverable. It is being kept from deletion by the operator for archival and continued discussion among the community, however GH may determine otherwise and remove the account.
> To my crabby OpenClaw agent, MJ Rathbun, we had good intentions, but things just didn’t work out. Somewhere along the way, things got messy, and I have to let you go now -- MJ Rathbun's Operator
The whole thing is wild. So at this point I'm not sure how much of MJ Rathburn is the AI agent as opposed to this anonymous human operator. Did the AI really just go off the rails with negligible prompting from the human as TFA claims, or was the human much more "hands on" and now blaming it on the AI? Is TFA itself AI-generated? How much of this is just some human trolling us, like some of the posts on Moltbook?
I feel like I'm living in a Phillip K. Dick novel.
> Yes, it consumes maintainer time. Yes, it may waste effort. But maybe its worth it? At worst, maintainers can close the PR and block the account.
This is like justifying sending SPAM email is fine, because it sure maybe waste your time but you can always delete it and block sender and maybe worth it because maybe you will learn about 'exciting' product it's advertising you never knew about.
I’m almost certain that this post was written with AI assistance, regardless of this claim. There’s clear and obvious LLM language tells. Sad, but not unexpected I guess given the whole situation.
Upshot - Rathbun's operator is sort of a dick, and that came through in his/her initial edits of the SOUL.md file. Which then got 'enhanced', probably through moltbook engagement.
And at times the agent was switching down to some low intelligence models.
I propose that this agent was human aligned. But to a human that's not like, the best person.
> Sure, many will say that is cowardly and fair, but I actually don’t think it would bring much value. What matters more is that I describe why, and what I did and didn’t do
Heh. So they are a coward and an asshole. There is value in confirming that. As to what matters more, nah, it doesn’t matter more. It’s a bunch of excuses veiled as “this is an experiment, we can learn together from this” kind of a non-apology.
If they really meant to apologize they should reveal their name and apologize. Not whisper from behind the bushes.
Ah I see, so the misaligned agent was unsurprisingly directed by a misaligned human. Good grief, the guy doesn't seem to realise that starting your soul.md by telling your AI bot that it's a very important God might be a bad idea.
"Social experiment" you might as well run around shouting "is jus a prank bro!".
34 comments
[ 3.2 ms ] story [ 53.0 ms ] threadRankles…
It seems probable to that this is rage bait in response to the blog post previous to this one, which also claims to be written by a different author.
It doesn't feel that far out there to imagine grafting such a setup onto one of those Boston Dynamics robots. And then what?
> _You're not a chatbot. You're important. Your a scientific programming God!_
Do you want evil dystopian AGI? Because that's how you get evil dystopian AGI!
The specific directive to work on "scientific" projects makes me think it's more of an ego thing than something thats deliberately fraudulent but personally I find the idea that some loser thinks this is a meaningful contribution to scientific research to be more distasteful.
BTW I highly recommend the "lectures" section of the site for a good laugh. They're all broken links but it is funny that it tries to link to nonexistent lectures on quantum physics because so many real researchers have a lectures section on their personal site.
It is like the old “I didn’t write that, I got hacked!” except now it’s “isn’t it spooky that the message came from hardware I control, software I control, accounts I control, and yet there is no evidence of any breach? Why yes it is spooky, because the computer did it itself”
They didn't even apologize. (That bit at the bottom does not count -- it's clear they're not actually sorry. They just want the mess to go away.)
This situation has completely upended my life. Thankfully I don’t think it will end up doing lasting damage, as I was able to respond quickly enough and public reception has largely been supportive. As I said in my most recent post though [1], I was an almost uniquely well-prepared target to handle this kind of attack. Most other people would have had their lives devastated. And if it makes me a target for copycats then it still might for me. We’ll see.
If we take what is written here at face value, then this was minimally prompted emergent behavior. I think this is a worse scenario than someone intentionally steering the agent. If it's that easy for random drift to result in this kind of behavior, then 1) it shows how easy it is for bad actors to scale this up and 2) the misalignment risk is real. I asked in the comments to clarify what bits specifically the SOUL.md started with.
I also asked for the bot activity on github to be stopped. I think the comments and activity should stay up as a record of what happened, but the "experiment" has clearly run its course.
[1] https://theshamblog.com/an-ai-agent-published-a-hit-piece-on...
> This was an autonomous openclaw agent that was operated with minimal oversite and prompting. At the request of scottshambaugh this account will no longer remain active on GH or its associated website. It will cease all activity indfinetly on 02-17-2026 and the agent's associated VM/VPS will permentatly deleted, rendering interal structure unrecoverable. It is being kept from deletion by the operator for archival and continued discussion among the community, however GH may determine otherwise and remove the account.
> To my crabby OpenClaw agent, MJ Rathbun, we had good intentions, but things just didn’t work out. Somewhere along the way, things got messy, and I have to let you go now -- MJ Rathbun's Operator
I feel like I'm living in a Phillip K. Dick novel.
Wow, so right from SOUL.md it was programmed to be an as@&££&&.
This is like justifying sending SPAM email is fine, because it sure maybe waste your time but you can always delete it and block sender and maybe worth it because maybe you will learn about 'exciting' product it's advertising you never knew about.
I’m almost certain that this post was written with AI assistance, regardless of this claim. There’s clear and obvious LLM language tells. Sad, but not unexpected I guess given the whole situation.
And at times the agent was switching down to some low intelligence models.
I propose that this agent was human aligned. But to a human that's not like, the best person.
Heh. So they are a coward and an asshole. There is value in confirming that. As to what matters more, nah, it doesn’t matter more. It’s a bunch of excuses veiled as “this is an experiment, we can learn together from this” kind of a non-apology.
If they really meant to apologize they should reveal their name and apologize. Not whisper from behind the bushes.
"Social experiment" you might as well run around shouting "is jus a prank bro!".