I like sunsets, long walks on the beach, functional programming, type systems, interpreters and compilers, and many other things which don't pay the bills.
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...
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?
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".
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.
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.
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.)
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"?)
I… yes. That’s a reasonable interpretation of the current state of the world even if I reject the “no choice” framing. But you’re the one calling his stance “easy”?
I guess I'll say it unhedged (since nobody else has yet): I agree with the author. I don't 100% align with his particular assignment of threats and externalities to AI, but I think overall it sucks and I don't want to use it. I think he's in a space that's getting overwhelmed by AI and AI output and I salute him putting his money where his mouth is. I wish him luck.
I'm sorry, but what on earth are you talking about? This guy appears to be a web designer. You think there's no competition in that space and no incentive to churn things out fast with AI? Agree or disagree but if he's honest his money is very much where his mouth is.
I am reasonably confident I am not aphantasiac in the sense described here. My apple is red with some nice green mottling, dust, etc etc. I can see my wife and children in detail in my "mind's eye". As a child people told me I had a "photographic memory", which I also think is sort of "fake" or semantic, but nonetheless: I can remember images I have seen with visual detail and I can imagine hypothetical objects "visually". I still think many self-diagnosed aphantasiacs are tripping on the semantics of wording in a self-report survey, because I could easily talk myself into the other label based on second-and-third-guessing "but what does it mean to _see_ something I'm imagining?" etc. If there is a group of people with an actual quantifiable brain difference, I suspect it is much smaller than the group of "all self-diagnosed aphantasiacs".
I'm not really sure what the hostility is about. That's an interesting study albeit small: 42 non-aphantasiacs and 18 self-diagnosed phantasiacs. If one really wanted to quibble I'm sure there are other possible confounding variables which could be involved in both pupil dilation and self-reported aphantasia. I'm not "blaming" neuroticism, I'm pointing out that these self-reporting surveys are subjective in a way that can be influenced by personality traits.
I'm very convinced this is what's going on. Qualia are super hard to convey with language! If I had to guess these self-reported instruments are also measuring, indirectly, conscientiousness (do I _really_ see it? it'd be wrong to lie on a survey!) as well as neuroticism (oh God what does that mean? do I _really_ see it? do other people _really_ see it? am I broken?).
You got it. You may well be right, but from my standpoint it's hard to see how you'd know if you weren't. Yeah, the models fill in vagueness with detail sourced from "what's likely and helpful". (In one of my particular domains, that's "use SHAP values for everything everywhere".) Yeah, that detail's increasingly unlikely to be "wrong" in a way that you could explicitly argue. Is it what an expert would come up with? Is it the best it could be? Is there something else that we would have done had we been forced to think it through? Will it ever be _surprising_?
I know, all irrelevant little philosophical roadbumps. The artifact satisfies the buyer. I think I'm going to have to step away from this industry pretty soon, at least for a while.
I've been thinking about the Blight and [spoilers!] the skroderiders a lot lately. If you have read Fred Saberhagen's Berserker stories, I have also been thinking a lot about "goodlife" when I see some of the things that get posted. If you haven't, you might like them.
I see a lot of speculation that this is somehow due to the recently-announced "watermarking", but that seems really unlikely to me, because the watermarking is about nudging per-token output probabilities or the sampler in a detectable way. That's not going to change the structure of every output to honest caveat this and verbing abstract nouns that. For the examples people are showing, and I've seen plenty of gibberish myself, you don't get there by perturbing probabilities from any sane starting point.
I don't know about the top people but Anthropic has always been pretty deeply linked with the effective altruism/AI risk/TESCREAL memeplex. IIRC one of OpenAI's people who's pretty prolific on Twitter weighed in a while back to the effect that Anthropic mostly sees themselves as an organization dedicated to serving Claude. Not serving like server but serving like... worship. I find them a little freaky to be honest.
No, it felt too invasive and creepy so I abandoned the idea. Or my lack of follow-through is a sign that I’m actually on the other side of the “war”. Reader decides! But suffice it to say that it was, to me, a significant trend especially around the “Claude Code moment” back in February/March is this year.
I think we’re on opposite sides but this is the first time I’ve seen someone in the wild share my suspicions that the LLM conflict is part of some vast ADHD vs autism hyperwar. At one point I was collecting posts I found where LLM boosters explicitly mentioned how it interacted with their ADHD.
I wondered the same but after running some previous entries through Pangram (and giving them the ol' Mark I Eyeball), I don't think this is satirical. How annoyingly ironic.
I'm a little confused in that this post is either (partially?) AI-generated or deliberately written to appear so (including to Pangram).
I'm a self-confessed AI hater and yet I agree with the "just write your thing" people on this one. I've watched this play out on Substack and I think there's a mix of a) people genuinely scared of what I believe to be very rare false positives from automatic detection, b) people who use AI (but claim otherwise) trying to use group (a) as cover to argue against automatic detection, and c) the hoi polloi throwing rotting fruit at anyone who uses an em dash.
You can safely ignore group (c); they aren't going to be your people even with a fully-attested human changelog witnessed by the Pope and encoded on the blockchain. Group (b) are basically grifters and also to be ignored. Group (a)'s concerns are valid but in the large I don't see a lot of "grey area" writing online; I see a lot of obvious slop. (Possibly the well-done AI assistance goes unnoticed by me, in which case - philosophical concerns aside - I can only think of the old XKCD comic about spam bots - mission accomplished? I don't think so, though.)
Back to the OP, I didn't really understand the "meet the enemy" section - did that experiment go anywhere? My results from trying to see if even frontier models can capture voice, style, philosophy from my own technical writing corpus have been terrible to laughable at best.
(Sorry, I'm in a crappy mood, but what on Earth are we supposed to take away from this? Everyone who disagrees with you is secretly an idiot, or worse, they're smart enough to know they're idiots but too proud to admit it?)
On a more helpful note, I think your "confusion" if honest can probably be resolved by realizing that "skeptics" are not a monolith.