Also, the reporting type news still needs someone to go out and get the news. Where do we think the news comes from? Sports reports, sure. 90% of it can just be pooped out by some LLM (the really good stuff will always belong to humans with insight into the sports; I don't even want to hear non-cyclists report on the Tour de France, let alone a device that's never actually played any sports at all). Some kinds of weather and financial reporting, maybe. Also some political reporting as well. But reporting involves people going places and asking questions.

And particularly as LLMs lower the difficulty of FAKING news, I think we'll want MORE people that were verifiably AT places. I've been floating the idea for years that Fox, NBC and CNN (or whomever) should be sending reporters to major events together to confirm that they actually happened. They can still report on them with whatever bias suits them, but together they can all confirm that all the things that were said on tape were actually heard by several different humans. It would be a selling point for all of them.

Yes, the Sears catalogue type of model will almost certainly disappear. And some types of conventional model will disappear.

But have you ever heard of the OKCupid study that said that people that have more polarizing looks tended to get more messages than people that everyone agrees are hot? It's not that I don't think that AI can replicate whoever is considered attractive at the time, it's that I think AI is a *trailing indicator* of attractiveness.

So AI has a good sense of what's attractive right now, but new people will crop up now and then and dazzle us with novelty--something different from everything we've been inundated with, and AI will move to that spot, but our quest for novelty won't be easily satisfied. I think just the existence of internet porn is evidence of that. Surely by this time, the permutations are mostly covered, but there's still a market for a new face doing all the things you saw the last person do.

I would say that most hotels ALREADY don't have a concierge is the point. I agree that they'll put a kiosk or robot there, but there was nothing there before.

14. Concierges

To the extent that these jobs exist at all, I don't think they're going away any time soon. Part of the benefit of a concierge is the prestige of having one, or to have a real person coordinate and organize things for you. I think more people might have access to a facsimile of a concierge now. No longer for fancy hotels and rich people only.

15. Political Scientists
16. News Analysts, Reporters, Journalists

Um, no. To be good at any of these things, you need to have lived in this world as a human and have real human concerns. AI can't show up at press conferences or do investigative journalism by itself, so I think these ones are majorly overblown. Will the big newspapers and cable news networks survive? Totally different question. I myself support a few independent journalists that do good work because they're more local and more specialized than the big Canadian newspapers are.

20. Hosts and Hostesses

What? Like, am I misunderstanding what they think these are? Someone that greets you at a restaurant or similar? That takes you to your table? You certainly don't need an AI to coordinate table assignment, the host(ess) is there to give a human touch to it rather than let you seat yourself.

29. Data Scientists

Anything with science in the name is off the table. At the very least, they need to be there to supervise the AI. But again, human concerns are what lead to human analysis. Generative AI still can't do new things or ask meaningful questions.

35. Geographers

Spoken like someone that thinks geography is just about maps. I have a friend that did a geography PhD on the sounds of a geographic region. Not just the ambient sounds of a forest or a city, but about the music and the culture. She recorded military planes landing with killed soldiers on them--obviously you can't tell that from the sound, but as part of the context, it's something that she felt was worth considering. Why would an AI ever think about that? Again: human concerns.

36. Models

Sort of. I think we're seeing this already, but I think there's always going to be a place for beautiful people that we know are real, beautiful people. Fashion and attractiveness move in cycles (see the cycles of something like jeans, or moustaches)--people love novelty. AI will only ever be able to respond to those things, because by the very nature of trying to make a thing happen, the counter-culture moves into the position of novelty and becomes the new hotness. AI generated perfection will become a dominant aesthetic for a while but then some other weird thing will pop up and they'll have to move to cover that after.

Anyway, some of these are probably sadly dead professions, they're not wrong. Translation is like the origination of LLMs, that's only going to get better and better. I can teach myself the basics of almost anything now with ChatGPT as long as I'm diligent and skeptical. Nobody likes telemarketers and nobody will be sad to see that go or switch over to AI, honestly.

I was pretty sure you were joking, but man, you can never really tell on the internet. :D

Almost certainly not. Buying Apple certified parts is indeed INSANELY expensive, but they're probably not taking much of a profit on it. While they obviously do break and need to be fixed now and then, iPhones tend to be relatively durable and people just like buying them. iOS 26 is still going to run on the iPhone 11, so I think it would be hard to make a case that this is just planned obsolescence or whatever. (I still have a working iPhone 11, in fact. Kept a decent case on it and it was great for years.)

I've been using ChatGPT to learn R and Python. I tell it up front to never write any code for me, to give me links to documentation, and then I ask it specific questions about syntax. I give it my code to look over once I'm done writing and testing it, to see if it has any suggestions that make my code more idiomatically correct. It's been working fine.

Like, you don't have to use ChatGPT to cheat if you actually want to learn. The problem is that these people don't care about knowing things. The system encourages good grades, not good understanding. The teachers don't NEED those problems answered correctly for their sake; they already know how to do those problems and find the answers.

This mode is meaningless, there's nothing new here. There's probably a few background prompts that are like my initial prompts that just say "avoid giving the answers, even when asked directly". As with so many things in the industry, it's just theatre.

These people definitely exist. I don't think it has much to do with their gender or orientation, though. It is remarkable how unhinged individuals can be, but it still doesn't speak to the broader categories that they exist in.

Based on this comment, I'm sure women are too.

I think we've all experienced this. At 30-ish, suddenly you can't drink as much, your neck and back start to get tweaked more often, you're not as powerful an athlete as you used to be (though sometimes you get better in other ways).

At 40 my eyesight started to go. I started wearing computer glasses. My aches and pains were worse, and I took so much longer to heal. I went on statins because even while I stayed athletic and ate pretty well, I just couldn't get my triglicerides any lower.

I'm 48 now, and I'll need to start wearing glasses all the time, soon. I don't recover as well from my bike rides, I have fewer great days. One bad night's sleep will haunt me for the whole week.

Better than the alternative. Supposedly people over 60 are just generally happier and more satisfied with life. I'm just trying to get to that part.

Billionaires are not workers; they may once have done work, but at the point where they're billionaires, their money is predominantly from OWNERSHIP not from INCOME. They aren't paid a salary (famously, some CEOs take $1 as their salary) that's meaningful, they aren't compensated by the hour. They're 'paid' in ownership and the ownership is what generates income. They make money off the work of other people.

Some CEOs are workers, no doubt, especially at smaller companies. They do some organizational and planning work. But once your wealth comes from owning rather than labouring, you're not a worker anymore. (Landlords aren't labourers either; by and large they do no work. To the extent that they work on upkeep, the actual JOB is being a maintenance worker, not a landlord. Landlords can and often do hire workers to maintain the property.)

Maybe you've seen this paper, maybe not:

"Recent math benchmarks for large language models (LLMs) such as MathArena indicate that state-of-the-art reasoning models achieve impressive performance on mathematical competitions like AIME, with the leading model, Gemini-2.5-Pro, achieving scores comparable to top human competitors. However, these benchmarks evaluate models solely based on final numerical answers, neglecting rigorous reasoning and proof generation which are essential for real-world mathematical tasks. To address this, we introduce the first comprehensive evaluation of full-solution reasoning for challenging mathematical problems. Using expert human annotators, we evaluated several state-of-the-art reasoning models on the six problems from the 2025 USAMO within hours of their release. Our results reveal that all tested models struggled significantly: only Gemini-2.5-Pro achieves a non-trivial score of 25%, while all other models achieve less than 5%. Through detailed analysis of reasoning traces, we identify the most common failure modes and find several unwanted artifacts arising from the optimization strategies employed during model training. Overall, our results suggest that current LLMs are inadequate for rigorous mathematical reasoning tasks, highlighting the need for substantial improvements in reasoning and proof generation capabilities."

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Farxiv.org%2Fabs%2F2503.219...

But you do make it clear that you think that regurgitation is sufficient in this case, and I won't argue with you; I'm no mathematician. Certainly, I believe that LLMs can be useful as long as we're not fooled into thinking that THEY'RE thinking.

Yeah, I think every generation thinks this.

But honestly, if you haven't found anything full of human spirit and culture, that's probably on you. There's still tonnes of original music made by indie bands out there. Go looking for modern punk/reggae/ska bands that are writing protest songs. Maybe explore some of the culture of other continents--I've heard that French African music is undergoing a bit of a renaissance (and there are plenty of European french people lowkey mad about it because they never like slang coming in from outside the country).

But even the pop stuff--yeah, it's generic. It always has been. That's what makes it pop music. But they've still picked up on the last 20 or 30 years of musical trends and integrated them into their songs. The tone and musicality of the songs has shifted slightly. It's not for me either, but it's not meaningfully better or worse than the pop of the past, tbh, it's just not MY music.

AI slop is so much worse, because in general, it's ONLY copying music from a single genre. Remember that musicians today probably listened to a bunch of stuff when they were young. If they're kids of GenX, they would've heard their parents' music growing up and then integrated some of that into their own work. You get weird crossings over of genres and sometimes something interesting pops out.

Anyway, AI slop can only ever be as interesting as the average of the current corpus of human music. It has no feelings, no experiences, no struggles. It will probably never write anything as hilariously bad as 'Friday' by Rebecca Black, which came back around and accidentally became the zeitgeist for a brief period, and now is actually weirdly nostalgic for a lot of people that experienced it at the time.

C++ is also largely changed by a committee. Honestly, the best thing that's happened to C++ in its whole lifetime is that the committee started ripping features off from other languages.

Gone are the days where a language can be designed, written and MAINTAINED by a single person. I don't know why we'd even want that, to be honest. Committees are fine as long as they get the job done.

On the one hand, this isn't in the job description, so...no.

On the other hand, I actually think it's useful for people that are programming systems that other people use to actually use the systems themselves in a production environment to see how they function. If you're a programmer writing software that people at the warehouse have to use, it SHOULD be part of your description to do that job for a few days a year to understand what the biggest problems are.

And also: no volunteering. If you spend any time doing this, they pay you whatever your hourly wage is + overtime. If you're a high-paid programmer and you do this, they pay you your programmer wage and compensate you for your time. It's such a drop in the bucket for them, there's literally no reason but greed not to.

Lastly: fuck Amazon and their shitty labour practices and horrendous (reportedly) work environment. I wouldn't work there on a bet.