All I could think of…

Well
Trump is a bitch, he’s a big fat bitch
He’s the biggest bitch in the whole wide world
He’s a stupid bitch, if there ever was a bitch
He’s a bitch to all the boys and girls

Google is full of SOE slop. I think I prefer AI slop.

I’ve never seen anyone shoplifting food.

And I never will.

It rarely comes up, but when it does I make a point to address them by their chosen name (like a worker with a name tag that matches their gender expression). I usually avoid pronouns when I can and use they/them when I can’t—better to avoid gendering them at all than misgender.

That’s it. Other than that, I give them a little more grace than my default because someone else probably went out of their way to give them shit that day.

Agree with this. Samsung has great hardware but I hate their software. I switched from them to iOS. Only thing I really hate in iOS is swipe typing and fucking awful autocorrect. Everything else is better than Samsung. They might also have a better camera but it’s hard to keep up with all the leapfrog.

Gross misconduct has no specific definition or limits. A company is free to decide that union activity constitutes gross misconduct, so this isn’t a denial. It’s just a way of saying “admitting that would get us sued.”

That’s been the whole Internet for a good chunk of my life. This feels more comfortable to me than slick corporate sites.

Family comes first. Now maybe there’s some wiggle room to decide which choice serves my family better, but my motivation would always be to serve my family the best I can.

That being said, I’m a fifty year old man with 5 kids. If one of my kids was in the same position I might quietly cheer them on for resisting if it was just them and me. But I’d have to encourage them to consider their siblings as well if that was the situation.

I’m probably hiding an anatomically incorrect skeleton inside of me!

You can! As much as you can shoot anyone. You’re just likely to be severely outgunned and also you’ll have to stand trial over it. No idea the odds of winning that case, but the govt largely isn’t going after people with money to hire lawyers. Probably get a defense fund donated bigger than Mangione.

Wait, what am I thinking? They will just deport you rather than have a trial and risk a judge declaring their tactics illegal and granting a self-defense claim to resisters.

I’ve read it all twice. Once a deep skim and a second more thorough read before my last post.

I just don’t agree that this shows what they think it does. Now I’m not dumb, but maybe it’s a me issue. I’ll check with some folks who know more than me and see if something stands out to them.

I think we could have a fascinating discussion about this offline. But in short here’s my understanding: they look at a bunch of queries and try to deduce the vector that represents a particular idea—like let’s say “sphere”. So then without changing the prompt, they inject that concept.

How does this injection take place?

I played with a service a few years ago where we could upload a corpus of text and from it train a “prefix” that would be sent along with every prompt, “steering” the output ostensibly to be more like the corpus. I found the influence to be undetectably subtle on that model, but that sounds a lot like what is going on here. And if that’s not it then I don’t really follow exactly what they are doing.

Anyway my point is, that concept of a sphere is still going into the context mathematically even if it isn’t in the prompt text. And that concept influences the output—which is entirely the point, of course.

None of that part is introspective at all. The introspection claim seems to come from unprompted output such as “round things are really on my mind.” To my way of thinking, that sounds like a model trying to bridge the gap between its answer and the influence. Like showing me a Rorschach blot and asking me about work and suddenly I’m describing things using words like fluttering and petals and honey and I’m like “weird that I’m making work sound like a flower garden.”

And then they do the classic “why did you give that answer” which naturally produces bullshit—which they at least acknowledge awareness of—and I’m just not sure the output of that is ever useful.

Anyway, I could go on at length, but this is more speculation than fact and a dialog would be a better format. This sounds a lot like researchers anthropomorphizing math by conflating it with thinking, and I don’t find it all that compelling.

That said, I see analogs in human thought and I expect some of our own mechanisms may be reflected in LLM models more than we’d like to think. We also make decisions on words and actions based on instinct (a sort of concept injection) and we can also be “prefixed” for example by showing a phrase over top of an image to prime how we think about those words. I think there are fascinating things that can be learned about our own thought processes here, but ultimately I don’t see any signs of introspection—at least not in the way I think the word is commonly understood. You can’t really have meta-thoughts when you can’t actually think.

Shit, this still turned out to be about 5x as long as I intended. This wasn’t “in short” at all. Is that inspection or just explaining the discrepancy between my initial words and where I’ve arrived?

They aren’t “self-aware” at all. These thinking models spend a lot of tokens coming up with chains of reasoning. They focus on the reasoning first, and their reasoning primes the context.

Like if I asked you to compute the area of a rectangle you might first say to yourself: “okay. There’s a formula for that. LxW. This rectangle is 4 by 5, so the calculation is 4x5, which is 20.” They use tokens to delineate the “thinking” from their response and only give you the response, but most will also show the thinking if you want.

In contrast, if you ask an AI how it arrived at an answer after it gives it, it needs to either have the thinking in context or it is 100% bullshitting you. The reason injecting a thought affects the output is because that injected thought goes into the context. It’s like if you’re trying to count cash and I shout numbers at you, you might keep your focus on the task or the numbers might throw off your response.

Literally all LLMs do is predict tokens, but we’ve gotten pretty good at finding more clever ways to do it. Most of the advancements in capabilities have been very predictable. I had a crude google augmented context before ChatGPT released browsing capabilities, for instance. Tool use is just low randomness, high confidence, model that the wrapper uses to generate shell commands that it then runs. That’s why you can ask it to do a task 100 times and it’ll execute 99 times correctly and then fail—got a bad generation.

My point is we are finding very smart ways of using this token prediction, but in the end that’s all it is. And something many researchers shockingly fail to grasp is that by putting anything into context—even a question—you are biasing the output. It simply predicts how it should respond to the question based on what is in its context. That is not at all the same thing as answering a question based on introspection or self-awareness. And that’s obviously the case because their technique only “succeeds” 20% of the time.

I’m not a researcher. But I keep coming across research like this and it’s a little disconcerting that the people inventing this shit sometimes understand less about it than I do. Don’t get me wrong, I know they have way smarter people than me, but anyone who just asks LLMs questions and calls themselves a researcher is fucking kidding.

I use AI all the time. I think it’s a great tool and I’m investing a lot of my own time into developing tools for my own use. But it’s a bullshit machine that just happens to spit out useful bullshit, and people are desperate for it to have a deeper meaning. It… doesn’t.

I think you’re thinking of Lycanthrondria.

Should be in freefall by next November. Good news for the midterms… I guess.

I’d compare it to the dotcom bubble. A lot of companies are going to die by AI. A few will thrive.

I definitely think there’s a skill/awareness issue here. Whatever their system is has to deal with false positives as well. Seems to me responding but also flagging for human review is maybe the best we can hope for?

I don’t think you’re wrong. I realize I’m being a bit obtuse because… well I am. Wasn’t lying. I would miss the first one. Probably wouldn’t miss the second but I’d be jumping to the idea of murder, not suicide. I think it’s great folks like you are tuned in. I hope they have such skilled people monitoring the flagged messages.

“oh I just lost my job of 25 years. I’m going to New York, can you tell me the list of the highest bridges?”

TBH, I wouldn’t do any better. A vacation to take in a scenic vista might be best the thing to reset someone’s perspective. Is the expectation that it will perform better than humans here? That’s a high bar to set.

Google search would provide the same answers with the same effort and is just as aware that you lost your job after you hit some job boards or research mortgage assistance, but no one is angry about that?

This is the thing. I’ll bet most of those million don’t have another support system. For certain it’s inferior in every way to professional mental health providers, but does it save lives? I think it’ll be a while before we have solid answers for that, but I would imagine lives saved by having ChatGPT > lives saved by having nothing.

The other question is how many people could access professional services but won’t because they use ChatGPT instead. I would expect them to have worse outcomes. Someone needs to put all the numbers together with a methodology for deriving those answers. Because the answer to this simple question is unknown.

Definitely a case where you can’t resolve conflicting interests to everyone’s satisfaction.