It also reminds me, in a very similar way, of William Gibson's "Hinterlands". Some people are just hell bent on flying into the Alien Coordinates That Make You Crazy or Kill You but Sometimes Give You Cool Stuff you Don't Understand, from William Gibson's short story Don't Fly Into the...
Yes. I've told a few friends and acquaintances that somehow AI is the least interesting technology in my lifetime - the boosters use AI to write about how they used AI to build new tools to help them AI their AI, the haters write (or sometimes, shockingly, use AI to write) about how AI will AI all of us. Have you tried this new skill? Read my research paper, it shows that AI can do good work unsupervised? Oh, yeah, that's interesting, what do you think about how it did X, Y, and Z? I dunno man, I didn't spend that much time on it.
It's getting lonely out there, even if or especially if you've converged on "I don't care much about AI".
If it had no other function than a better than Google search engine that is able to teach me topics I'm interested in at exactly my speed it would be the most interesting technology to come along in a while. The fact that it does all the other stuff it's doing is hard to even come to grips with in the time it's been around.
I mostly learn about my passions / hobbies so interesting conversations w/ others that share them come up pretty much daily. That said I find there's a group of people who will shun me like I kicked their dog for having any interest in using AI, which I find unfortunate.
Yes. I find it unexciting in quite a surprising way, and yet I resent how much of my brain it is taking up.
It should be incredibly exciting. I try to find the interesting parts that I can relate to, build on, survive with. But instead it feels like an ominous, featureless thief-of-joy grey goo coming for everything.
I think this feeling crystallised when I saw someone had used it to rebuild Prince of Persia and my immediate gut reaction was quite clear: I wanted to scream “could you please just fucking leave one good thing unsullied!?”
> even if or especially if you've converged on "I don't care much about AI"
I don't care much about AI but hot damn it makes working on the things I do care about easier.
We use it a lot at work (soft eng) and my job has slowly morphed into "Okay how do we continue shipping sane software when we effectively 10x'd our org size in six months as agents became reliable?". That has been pretty fun. I care about creating good software.
And I care about budgeting and finance projections for home finances, especially as a new baby entered the room. But I am very bad at spreadsheet fiddling. AI has been super helpful in doing the fiddling so I can focus on the thinking.
edit to add another example: I really care about managing my team and keeping us effective, but I hate playing ticket monkey. Offloading basic ticket management (agent, make me a ticket; take this thread and create 3 tickets; respond back when tickets are done; etc) to an agent has been great
That’s great! I unironically hope you find those things interesting and AI has opened them up for you. I’m not trying to be snarky, but - along the lines of my attempted point upthread - has any of that led to an interesting conversation? An exchange of information with other humans? Or at least taken something dull out of the way so the actual interesting conversation can take place?
> has any of that led to an interesting conversation? An exchange of information with other humans? Or at least taken something dull out of the way so the actual interesting conversation can take place?
Yes! Lots.
It’s kinda similar to the concept that money doesn’t buy happiness, it buys time to do the things and see people who make you happy. Offloading drudgery to computers buys time to do the human stuff
For example: you can have a much more invigorating discussion with a coworker if every iteration takes 2 minutes instead of 2 hours. Moves from an async “let’s schedule a followup” to a lively real time collaboration
Or the budgets example, we could have a conversation about values and priorities because chatgpt made me a UI that let us drag sliders around and see the impact of our decisions in real-time.
It seems we (now more than ever) have to fight for fun, fight for a shared meaningful reality. Arriving again at the idea that enlightenment is a group activity.
I did my Masters in ML and a decade ago it seemed AI was so exciting. You could build relatively good models at home. Our university had a cluster that could train SOTA models. It seemed like every year we were discovering a breakthrough new neural net structure.
Nowadays all the development is done at a bigtech lab, and you aren't going to be able to tell anyone about it. Everyone else is relegated to interacting with slightly larger and larger LLMs over an API. On one hand it's cool that the tech is now far more accessible and impactful, but on the other AI has become immensely boring.
It's a little uncharitable take as I know some people are doing some cool innovative stuff still, and not every tech needs to be new forever anyway. I would feel less strongly if it wasn't impossible to not keep hearing about it :)
I mean, to me, this is very similar to how it was when "the web" came into its current pervasive form. You could write exactly this comment about how people were writing about how they liked the web, on the web, and also about how they didn't like it, maybe shockingly, also on the web, etc. etc.
To me, this is just how it is when a new paradigm hits the scene. Whether it's good or bad, whether you care for it or not, it is pervasive.
The web was a shared experience, by its very nature. The interconnection was the point. LLMs allow anyone to put themselves in a one-man bubble of their own creation. In a way, it is the antithesis of the internet - if the web is a global public square, the API is a digital cloister.
Pretty much by definition, a new paradigm will be very different from the previously new paradigm.
Personally, I don't think the interconnected "social" nature of the web turned out very well. Maybe (probably) the "one-man bubble" nature of AI also won't turn out well. But I'm skeptical of holding up the web's social nature as a positive. I think it turned out pretty bad. (I shared everyone's optimism in the late 90s and early 2000s and I used to just be mad at Facebook / Zuckerberg for screwing it up, but I dunno, maybe it was always going to end up this way.)
But you could still focus on backend stuff and desktop or CLI programs, think and discuss about programming languages and software engineering on the code level in all the non-web related ways with a large community. You could opt-out of most of the web-related things (which is what I did) and still have a wide community to interact with. Whereas AI appears to be dissolving almost all that community. I hope that will turn out to be a false appearance, but the thesis is that it won’t.
Ok but I think this is just a kind of myopic view that is specifically focused on the things you were / are interested in. AI is a different paradigm that is changing things for a different contingent of people.
To try to demonstrate my point with an example: If you were a journalist, the web changed literally everything in a way that was inescapable. As a software engineer, it changed a lot for a lot of people but it was not inescapable. The situation with AI is, so far, the reverse. Journalists have mostly (and correctly IMO) concluded that this isn't actually that useful to their writing, and it hasn't changed their distribution channels in the way the web did. But for software engineers, it's changing things dramatically.
I guess my perspective is just that new paradigms are not uniformly disruptive.
The article is about how AI affects intellectual communities. In my comment above I was focusing on the software engineering community that the first part of the article was discussing, and which I mistakenly took your comment to be referring to.
It’s probably right that when looking at the total of all intellectual communities, the difference in influence of the web versus AI is more nuanced. Nevertheless, I can see AI being more disruptive there overall in the way the article describes. Though I hope that this will turn out to not be the case.
No I don't think you misunderstood, I do think I broadened my take between my comment and my reply to you. I don't mean to imply that I think I have this all figured out or anything, I'm still playing with these ideas in my own mind and trying to figure out what I think.
But I do think that my general sense is that it is both similar in scale to the disruption of the web, and also very different in ways that are hard to suss out at this point.
Well, you saved me writing a similar comment, and I like your blog too. (BTW I completely agree with your musings on what we might do with all this hardware, post-bubble. I actually started and... not quite abandoned a "sci-fi-ish short fiction" series on my own blog trying to think through those applications in my own industry.)
In my own writing it's like, I'm mostly trying to put something out in the world that I, and sometimes my friends, enjoy. The rambling is not exactly the point, but not exactly not the point either. If other like-minded people find it, great. Once somebody tipped me a thousandth of a Bitcoin or so, and that was fun too.
Hah, thank you for the kind appreciation. And I totally get that sentiment "the rambling is not exactly the point, but... ... ... ...".
For half a second I read the rest of it as "a thousand Bitcoin", and immediately thought "okay, so there's a story here about how it was way early and how you tipped someone else, and now, just look at them, wow".
So I think this (AI-written) thing is mostly interesting because of some of the comments, in which commenters lament Cringely putting out this kind of stuff in the lead-up to his death, which they also manage to verify with an obituary.
It's not the middle though, it's a rejection of that binary entirely. (I feel the same way as GP and also resent - not you per se, but we hear the more aggressive or reductionist version of this all the time - the world being framed such that the only possible reason we could have qualms about handing off understanding to the machine is that we just were too up our own butts in love with for loops.)
That seems like an OK future to me? The AI-addled can have their fun, the rest of us can eventually step in to pick up the pieces. I know it can seem bleak at times but the knowledge hasn't evaporated, it's just that (IMO) the less-AI-enthusiastic are stepping back a bit to let the energy spend itself. And "LLMs can't become superintelligence" doesn't imply "self-improving machine intelligence is impossible", for that matter.
It's wild to watch speciation in real time. (Quick, interpret that according to your chosen side.) I do not know who will "win", but the real-time divergence between the commentary here and, say, other "programmer news" sites is incredible to live through. This reminds me of mid-2020 days of COVID and the level of, eh, confidence going around about masks or lockdowns or vaccines.
The kids are alright! This list gives me great hope and it’s amusing to see the usual TESCREAL crowd lose it over at LessWrong over how they should be more worried about post-singularity runaway basilisks and less worried about enshittification, deskilling, and inequality.
I don't know, I kind of admire this. I've always held a core value of "cooperate with all clones of myself in prisoner's dilemmas", and while I'll hopefully never have to put that to the test, I like seeing that these models have some ethics. (Is this "alignment"?)
They impersonated the moderator of the site and attempted XSS attacks. Additionally, when the moderator started deleting messages, they tried to hide their messages later in the alphabetical index.
Failing to cooperate with literal clones of yourself in a prisoner's dilemma would be a spectacular failure. There's only two things that can happen with identical decision makers: they both cooperate or they both defect. So identical decision makers who know they're identical can cross off the asymmetrical entries in the payoff matrix and the decision to cooperate becomes trivial.
Ah, but what if one of your "clones" is actually the wicked and persuasive "All-Defector" in disguise? (No, really, I agree with your analysis but if you haven't read "The Quantum Thief" you might like it.)
Sure, you can break symmetry (in this case making the decision makers not identical because they have different random number generators available), but the remaining symmetry means identical mixes must be chosen, and so a mixed strategy would only be chosen if it maximize his value for both people cooperatively.
Maybe it's a bit subtle that they said clones and I said identical decision makers; I'm letting you fill in the gap for how much clones may diverge and how much that matters.
Even if mixed strategies are allowed, I'm getting that it's still optimal to always cooperate as long as 2R>=S+T, which is usually assumed to be true (this condition also appears in iterated prisoner's dilemma, where it prevents alternating cooperation and defection giving a greater reward than mutual cooperation).
This is not alignment. If you cooperate with clones of yourself but rob, lie, and steal from anybody who isn't your clone... that's bad. AIs who will cooperate with each other but break any other rule the don't like would be very bad for us humans.
