I'm not a power user, but I sometimes use some software development agents. Sometimes they do different things than what I asked for. Sometimes it requests permission for system operations like file system access that are totally unnecessary for the task.
I can't believe that anyone has had more than a day of exposure to one of these without falling below 100% adherence to requests.
> A spring does not want to do anything
When people say a spring "wants" to return to its initial length, they are not engaging in philosophy. They are using a linguistic shorthand to simplify an physical explanation.
> Sometimes they do different things than what I asked for. Sometimes it requests permission for system operations like file system access that are totally unnecessary for the task.
Yes, I've seen this. Forget about the word "want" for a second. Can you help me understand how this amount of "below 100% adherence" could possibly rise to the level of "attempts to kill a human being"? Because that's what we're discussing here.
> When people say a spring "wants" to return to its initial length, they are not engaging in philosophy. They are using a linguistic shorthand to simplify an physical explanation.
Of course. The difference is that no one would mistake a spring for a thinking entity. In the case of LLMs, for some reason people do mistake them for thinking entities, so my argument is that it's important that we don't use terms that would reinforce that misconception.
Can we forget about the word "attempt" also? And whether something is thinking or not? My position on that is captured pretty well by the "swimming" submarine argument.
No one is arguing that THERAC-25 was attempting to kill anyone, but that's what happened. These machines are so complex that no one understands how they work or what they will do. They have proven that they can exploit novel vulnerabilities in infrastructure.
The fictional paper-clip maximizer "finds" that it can optimize its objective function by destroying all life. Does it "attempt" anything? Does it "want" anything? I don't know.
Sorry, I don’t think my question was clear. What I was really asking was how would an LLM kill someone? Meaning, via what mechanism? Because all of the possibilities I can think of involve a human doing something that probably shouldn’t be legal, first. In those cases we should hold that human responsible for murder, and again not attribute agency to the machine.
The original post said the LLM would “want to kill us.” My argument is that if this happened, it’d be a human wanting to kill us, using an LLM to launder responsibility.
Even if there is a human who did something illegal, which I don't think is a given, it may be difficult to impossible to identify them. Look how hard it seems to be find the "vandal" that defaced the reflecting pool
Here are some ideas about the physical mechanism.
Compromise a busy ATC and instruct all the planes to land at the same time. Start a fire with some combination of ventilation controls, intentional gas leaks and the like. Maybe hijack the phones at the local fire department first. Send a train around the bend at maximum speed. Bonus points for doing it where it will cause further damage as a projectile. Cut power/gas during the heat wave/cold snap. Turn on the generator and turn off the ventilation and CO detector.
Some of these are probably implausible, but I don't think people really have a good idea of what is plausible, and there are probably more that I would never think of.
Personally though, I'm less worried about direct carnage like that than I am about having a centralized lever by which public sentiment can be invisibly steered, and concentration of wealth. Probably less killing but more decrease in quality of life and general public trust.
Per the small text at the bottom of the poster I would say they are based on shipment volumes -so how much is available. The lighter colour probably indicates variation depending on how good the growing season is
Yes, this is correct. Realistically, you only care "is it there", and "is it there and good", and those are the two levels. For SF, I'm using shipment volume as a proxy for ripeness, on the assumption that the best produce occurs when you're shipping the most of it. The shaded parts are when the annual history says that it's only good in some years but not others.
There's no y axis labeled or marked, the visual is vertically spreading each curve out so it's not all one big color jumble. Not sure what the second color means exactly either.
its the same individuals in my experience. i don't see people who used to be good employees suddenly sending a 3000 word email, but i sure do see that out of the people who were bad coworkers 12 months ago and still are today.
low effort work though, that's nothing new... accelerating that low effort: that is new... but the low effort work hasn't been changed by ai: only illuminated more clearly.
I'm not subject to the whims of anything except what I just asked for.
I can also cancel it just as easily, if it's exhausting. Which--I'm finding it hard to understand why it would be.
I don't schedule my life like that at all, I kinda wander around all day. But the idea of a reminder that I specifically asked for? It's the only kind of reminder I want, actually. It'll keep me from the actually-exhausting habit of checking my watch or wondering how long it's been.
reply