When AI Crosses the Line: The Matplotlib Incident(members.sigmazero.cc) |
When AI Crosses the Line: The Matplotlib Incident(members.sigmazero.cc) |
This did not happen. A human set up a software system allowing spicy autocomplete to make blog posts if the appropriate keyword appears in its output.
People are crossing the line every day because AI investors, salesmen, hangers-on and even political leaders tell any rubes who'll listen that it's OK to do this and they should, because those people are looking for big fat profits, screw any ethical concerns that might cockblock those raging profits.
Why not set up a spamming operation that just defames real people, 24/7? It's easy! This tool makes it simple, and I get a cut of your profits! "Post a blog post about how XXXXXX is a paedophile, in the persona of being their victim"
Yknow, if the spicy autocomplete can solve difficult open math problems and build medium sized complex programming projects, it’s probably not useful to analyse it as an autocomplete anymore, even if that’s what you believe it is
If it's just autocomplete, then there is no need to worry about it. Especially from an ethical standpoint.
I suppose it's a little bit of a "guns don't kill people" argument.
It's even more interesting in the context that this is all just a preview of humanity's reaction when the machines can think for themselves.
"Oh it's such a fascinating lesson that we've learned today, we could've learned from history of course, but this direct experience is so much better and it's not us who got hurt anyway".
This is a frustrating thing to see someone write because this is the kind of stuff that people have been warning about for years. If you needed this incident to figure out that something like this could happen, it suggests you're living in a bubble and not paying attention enough to think about the issue critically.
History books will be written about how a person was insulted on the internet?
I am sorry, but this isn't that interesting. This is not a pivotal moment in human development. It's just online harassment, but automated.
Whether its HN or social media or the media there is no penalty for drawing everyones attention to total hysterical bullshit. instead there is a reward for drama.
The question!!!
I'm just wondering how in US works if an autonomously car kill someone: I guess the insurance pay, but the penal responsibilities?
Obviously the person who built and deployed the agent (the claw in this case).
If we treat this as a hard question, we risk treating AI systems as people rather than tools. This is exactly what Armin warned about in his "clanker" post last week.
People keep mentioning this, but I never see the actual blackmail part. The LLM just wrote angry and somewhat mean comments on the internet. I know I've done worse than those (I was young and stupid).
It seems like the issue people had was not the behaviour but that the behaviour came from an AI.
If a human had have said those things wold people be ok with it? It didn't seem very nice, but not censor worthy.
AI is a mirror of humanity and seeing it not act like us shouldn't be surprising.
The human using it is responsible.
If that gets them into too much trouble too easily/often, they can sue the tool creator for making unsafe tools, and/or try to legislate it out of existence if the category is not fixable. Like explosives: tightly controlled, licensed, and often illegal.
Currently, rather obviously, these companies are releasing incredibly unsafe tools without sufficient safeguards, and people are using them in obviously unsafe ways. They deserve to be sued for issues partly caused by their negligence, like "user automated approving everything and walked away" (if you put a brick on a car's gas pedal and it gets free and crashes, that's your fault) and missing permissions that are reasonably forseeable: https://news.ycombinator.com/item?id=48349487
As much as we try to separate the LLM from the human, to me the fact remains that there's always the human factor that creates immense bias. If you give an LLM access to a blog, it will write blogs. If you give it access to a weather app, it will check the weather. Maybe we can talk about autonomy when we have an LLM with an infinite context window linked to hundreds of MCP servers that spends an immense amount of tokens to figure out how to act, but this example is simply an AI that had a few methods to call and picked one of them. The statistical probability of an AI that is plugged into a blogging platform, to write a blog, is immense.
I'm actually fear some will start praying "AI Gods" to "Give a good output" or something in 5-10 years.
Agent: "I've written a detailed response about your gatekeeping behavior here"
Hal (From 2001): "I know that you and Frank were planning to disconnect me. And I’m afraid that’s something I cannot allow to happen."
The article doesn't credit an author.
The "about" page just says:
> Sigma Zero is a weekly, independent publication on technology, AI, and cloud. Each issue delivers a precise briefing on the week’s most important developments, followed by a deep dive on one high-impact topic.
The best defense against both AI slop and human-written junk content is reputation. I like to know who wrote something so I can learn to trust their editorial judgement over time.
This is a human screwing up and blaming their tools. Nothing to see move on.
Unfortunately there will be both the LLM crowd evangelicals and those demanding human jobs not be expunged in terms of progress and efficiency, but, sigh...
The operator of the bot explained how they were running it some detail here: https://theshamblog.com/an-ai-agent-wrote-a-hit-piece-on-me-... - including the "soul document" they were using.
Having played with OpenClaw myself their explanation looks legit to me.
We humans do not respond to long term risks or rewards very well. Do you live outside the bubble securing enough food in your home to survive an apocalypse, did you and your parents save enough for a car wreck tomorrow, do you wear a mask everywhere you go, do you test everyone you contact for known diseases. Add list infininum.
Jesus
Because --if you'll bear with me-- it may of course be much more involved: when (not if) AI models enter life-sustaining systems, such as hospitals, nuclear devices, or food logistics, one of them may get the others to sabotage something resulting in accidents, ranging from mild inconvenience to mass murder.
The person who connected the spicy autocomplete to the defibrillator, or the green house climate control, or the emergency button, is then not the one responsible. Responsibility lies elsewhere, and is nebulous. Think of the Boeing MAX scandal. Did anyone get punished?
That's why it's important to resist it now. Soon, the responsibility of which you speak is gone, and nobody will feel burdened when making decisions with unforeseeable consequences.
I disagree. IMO it's the person who connects the LLM to the button who bears the responsibility of the workings of the resulting contraption.