Agreed but at some point I am forced to work “at gunpoint” because I have a wife and kids who need a house and food and cars. I’m jealous of anyone in a position to simply quit.

I work for a company that works for another company in the hospitality industry. The software system is being updated (in part of a much broader system change) to no longer allow non-binary or unspecified gender. We aren’t writing that part, but have to support it. I consider it a shortsighted and cruel change. But I’ve also spent a 7 of the last 30 months looking for work. I’m over fifty and I’m currently trying to build my retirement savings back up from zero after that.

I’m not walking away just because of this change. Instead I’m making sure our software is easy to change back when world is ready for that once again. That’s the best I can do, and I’ve worked for companies engaged in much greater evil.

When I got hired on a contract for Uline I’d never heard of them. Then I found out that are huge contributors to the Republican Party and I was glad when they decided to replace me on that contract, but I couldn’t just walk away. That was the most morally conflicted I’ve ever been at a job. But it gave my family the means to thrive, and that is my first goal.

Not OP, but…

It’s not always perfect, but it’s good for getting a tldr to see if maybe something is worth reading further. As for translations, it’s something AI is rather decent at. And if I go from understanding 0% to 95%, really only missing some cultural context about why a certain phrase might mean something different from face value, that’s a win.

You can do a lot with AI where the cost of it not being exactly right is essentially zero. Plus, it’s not like humans have a great track record for accuracy, come to think of it. It comes down to being skeptical about it like you would any other source.

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.

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.