ArcHound 7 hours ago

To give points to the girlfriend - you can still derive some pleasure from manually working on the craft while recognizing you're not able to compete with the industry. You'd have to accept loss of the status, career and probably other things. But it is possible to do programming as a hobby. I know I do.

To offer something positive/constructive - there's plenty of respect, pats on the head and fun to be had on a local level. I thoroughly enjoy speaking at our cozy little python beer meetup, where I can see that my ideas and approaches are interesting and respected by people I know, like and respect. It really helps with the mental state - I can then go back to work to do the needful all day. It was easy to do this on the global scale, well, now it's gone.

Another issue then is how to get a good career. To that I don't have the answer, other than to find something valuable to offer. The value of creating software dropped sharply. It might go back up one day, but we're deep in the pit today. Seems that's not your immediate problem and that's also good.

In the corner of my heart I hope that AI will collapse somehow by itself. Because I worry what might happen when you displace income for so many people with no path to improve their situation.

  • noduerme 6 hours ago

    My girlfriend will appreciate your credit, I'm sure lol. I used her as too much of a foil in this piece.[0]

    And on a local level; it isn't about status for me, it's really about struggling to do something nice and feeling a sense of completion. So yes, that could work...

    I hope it will collapse too, buddy. Not for my own career, but like you said, because there's going to be no path for so many people. I feel like it wants to collapse, like an NFT. But then... a lot of things would have to go very right or very wrong.

    [0] https://en.wikipedia.org/wiki/Foil_(narrative)

    • ArcHound 5 hours ago

      Yep, it works wonders. Since the local scene is smaller, you get to know people well.

      Similar idea is a "hacker space" where people tinker with electronics together.

      Cheers mate and good luck.

rogual 12 hours ago

This is an excellent article, and puts into words what I too have been feeling.

Although I will say, the article is kind of a counterpoint to itself — it was clearly not written by AI, it's clearly the work of an interesting, intelligent person. The difference between this and what Claude would have written is night and day.

  • noduerme 12 hours ago

    Thanks. It was written by me in one outburst. Then it turned out to be too long to post on HN, so I started a substack account, which is something I've avoided doing for.... (how long has substack been around?) Idk. Maybe I'll write more. I like just reading star codex and so on, don't really write much, don't have a social media presence, except responses here. I got hit with le feu follet or something last night. That and six vodka sodas.

guitarlimeo 11 hours ago

Thanks, you managed to sum up my feelings pretty well. Good to know our kind are not alone in this crisis at least. Sharing this with others too.

_trampeltier 14 hours ago

I feel you. That's what I felt when I was young and QBasic was my world. Then came the internet and Linux. Everything was just there, you just needed to download it.

  • noduerme 12 hours ago

    Like... yes but it was still a challenge. Trading floppy disks before that, knowing when to call a BBS, then learning where to telnet or gopher, heck even learning how to use a search engine to find answers, there was an economy of what you knew which was tradable or worth something in exchange for what other people knew, and helped you and them both become better at the same time at solving problems.

    What I think we're waking up to now is a world in which everything we knew has been scraped, consolidated, and packaged for sale to the people and companies who formerly valued us for our expertise. I don't know if this is too extreme or melodramatic, but it kinda feels like having spent a whole life collecting rare band merch, shirts or something, and then your house is burglarized, and now you're walking down the street seeing someone wearing one of your shirts who bought it off Ebay. Only you can't tell if it's even yours or just a copy. Whatever effort you put in has been taken, you understand that, but you're not even sure exactly how. You know that they don't value it or respect it the way you do, but then they get into a car and drive off. I was going to say "with your ex-girlfriend" but that's just too much ;)

herdrick 14 hours ago

This is a well-written explanation of how many of us are feeling. But if you take on a hard enough thing, one that you think you + AI could only barely pull off, wouldn't that be worth doing? The intellectually difficult software projects still need a smarter human operating the AIs than do the easier ones.

Dream bigger is what I'm saying. But I'm not saying that's easy.

  • noduerme 13 hours ago

    Thank you for the positive attitude. I need more of that ;) You're kinda right, that's a case you could strive for, finding the edge of what can be done now is still interesting. But two problems:

    1. That really requires me to shell out money to a service. Or a ton of hardware. I was a graphic designer, barrista, party photoshopper and a cab driver until 2004, as I learned to write code on the job. I built my business, savings, even bought a house owning one Macbook Air at a time, every five years or so, and not needing anything other than my wits and a command line, Eclipse, or VSCode. I never paid anyone to assist writing code for me. Paying Anthropic to work on an edgy project is a different kind of risk than simply risking your time building an experimental project... and then you're dependent on someone else, which I never liked.

    2. What I've been working on, and the only things I'm inclined to work on, are a different type of project; it needs to be hushed and secret, closed source, and [edit] not something facing the public. So either for personally quietly making money or for personal artistic satisfaction. Even if I push my own code & AI together to the brink for a public game or something, it will be stolen within seconds. Not only my time, but my investment in AI tokens will go to waste, because someone else always has more compute and more money to do that; there is no moat possible anymore by being the first to come up with an idea.

    So with that economic calculus, I'm spending time and resources on code to project the stock market and baseball games, instead of writing fun stuff that other people would enjoy. And that's a real bummer for me.

    But again, I'm lucky, because I already earned a lot making things I love, and I have safe coding jobs still available.

    • guitarlimeo 11 hours ago

      I've found myself thinking of the exact same things. Second one especially, and it's true that if your piece of work relies on a simple (but great) idea that can be copied it should be developed in secret.

      But re: both, aren't they just convenient truths that you can tell yourself to stop trying? Like this guy Greg Guevara says in his presentation [1], "Fame is evil, yeah ok cool okay. Success is evil, how convenient". You are absolutely logically correct in both of your assessments, but by following them you're limiting yourself and letting yourself to be directed to the bottle or to GTAVI.

      Buuuut, I'm not sure myself either. It's like I can follow these same set of principles as you do, be principled, or choose not to and be a bit hypocritical and be able to engage with the crooked world by doing so. I've started to dabble into the direction of doing big creative pieces with AI. I don't like it, I don't feel proud of my work, I feel like a director, not a craftsman. I've not yet gotten over carrying this dissonance in my head, but maybe it will get better.

      [1] https://youtu.be/rZ6QJDqciac?t=275

      • noduerme 11 hours ago

        You said a lot, man. I don't think it's unethical to do big creative projects with AI. It's less fun, and maybe less profitable in this environment, but it's still something creative. I was really talking about projects where, y'know, you have an AI write customized spam emails to drum up business for an AI-driven website. Or lots of worse things. AI advertising is about where I draw the line, personally, but even that isn't pure evil, it's just... not to sound like a dick, I just think it's undignified. You know what I'm saying?

        Making some cool creative project with or without AI is still great for your soul, and maybe that's just what I need to do, regardless of how it works out.

        re: Baseball, well. It's the most number-driven and mathematical sport ever devised; nothing else comes close. And still, it's fucking baseball. One bad hop, one bad throw, everything goes on its head. My software isn't AI-based although I had Claude do some work on it to help chart and track things. Toward the middle of the last season I started to run into guys in sportsbooks who have had Claude build their own custom site for them, tracking wind speed and days off and all kinds of stuff. I'm interested to see whether this actually increases or decreases my edge. Wait, "interested" is underplaying it. All I want is for baseball to start again lol. My biggest fear is that AI will eat the world before March.

        [edit] also, just to say thanks, about your comment on the bottle and gta6. I may just be projecting my own spiral... like, sometimes I wonder, without AI would I still be feeling like there was no point in sitting down and coding new stuff? And maybe that's it, maybe I just happen to be having some kinda burnout. But it doesn't feel like I want to stop, I just want my art back.

        • guitarlimeo 7 hours ago

          Sorry for misunderstanding you a bit then.

          > I just think it's undignified. You know what I'm saying?

          Totally.

          > without AI would I still be feeling like there was no point in sitting down and coding new stuff? And maybe that's it, maybe I just happen to be having some kinda burnout. But it doesn't feel like I want to stop, I just want my art back.

          I understand you completely. I've been asking the same questions.

          My whole life I've had role models who are experts at their craft, and now AI can do their craft as well or even better than them. For example a recent one was Billy Basso for creating Animal Well with a custom engine completely on his own. A tremendous effort of 7 years, a magnificent show of skill and craftmanship. But if he did it today, I feel like the general sentiment would be like "if he used AI he would've got it done in 2". There's no respect for the journey anymore. It's like, anyone can get a helicopter ride to the Mount Everest.

          Same goes for music. I honestly can't know who composes without AI anymore, especially newer artists. I have to just trust the artists or better some people who vouch for them. But mainly I just listen to artists from pre-AI era.

          And last fucking depressing thought I had was regarding standup - there's no way to prove that people aren't using AI to come up with the jokes. Granted, coming up with good jokes is only half of the work, performing them well is hard, but still, you can't ever anymore be certain that someone has original material. And yeah AI mostly sucks with jokes, but last week when I tried there were couple good ones for every 10 shitty ones. Why wouldn't comedians just grind that to come up with materials?

          I hate it. I want my art back also. And what I mean by that, I want that it takes effort and sacrifice to do art.

          • noduerme 6 hours ago

            Hah. I've been working on shitty sounding music recordings lately, I call them the drunk tapes. One take per instrument, one night, three songs on piano and guitar, slide, vocals, harmonica, accordion. Nothing is perfect. Everything is off. But it sounds real.

            And on the flipside, a customer asked me to rewrite their credit card management system. I spent a week making detailed, branded cards that look like real credit cards for their brand for every customer's card, not random but based on colors (last digit of the card) and graphics (second to last), merging and de-duplicating along the way. Looks fantastic. They were very happy with it. I could have used AI, but I needed to just design and write code. I would've charged the same anyway.

            There is, or there will be, a thirst for the genuine artifact. Once every website looks like Claude designed it, and every new music artist is indistinguishable, that thirst will be unquenchable. But I don't know how long it will take, for people to learn to see through the all the tricks; and everything we do, remember, the tricks are copying us and learning as we go.

District5524 13 hours ago

Great article, thank you! So many things resonate, although many things in our lives probably couldn't be farther away. Same feelings with D&D - although my parents were a lot more permissive than that, and it was a great counterweight to early binge-gaming until mid90s, being afraid of addiction and being a workaholic at the same time, fear of loss of meaning to life without being a workaholic. And suddenly, loss of meaning in prioritising some practical knowledge over everything else due to (incoming) capabilities of AI, seeing myself as opposed to people who prefer to manage other people instead, i.e. push them around and make use of them to their obvious detriment. (My distant hope know is that perhaps now teaching other people by people will be valued more in the longer term, act as an interpreter of knowledge to others... Perhaps there some human effort is still appreciated.)

  • noduerme 13 hours ago

    Huh, wow. You raised two interesting features of my own psychology I didn't even think about, and that might be worth writing about. Workaholism and being in opposition to managers who don't know how or don't get their hands dirty.

    I'm a workaholic; otherwise I'm an alcoholic. (Sometimes both ;) The need to avoid slipping into a bad addiction is what drives my "good" addiction.

    Then too, I've never understood managers. I'm not one who takes the view of seeing them as merely parasitical; they serve a capitalist/market function. Although I'd never want to be one. But the best ones are so unusual and special that they stand out in my memory. Like Rashad, a manager at a restaurant where I waited tables in 2001. Most other managers stayed down in the office, or just gave people crap. He looked around the floor and immediately saw who was overloaded, who was in trouble; he could target the situation and step in to fix it, wait the table himself if he had to. He was like a waiter and a chef and a busboy and a bartender, who had worked in every role and knew the exact details to take care of.

    But that's not what people do who "manage AI". Not at all. They seem almost proud of not needing to know how things work, as if it's a big "fuck you" to everyone who spent a lifetime learning how things work.

    I saw that you're a lawyer. I have some clients in my coding business who are lawyers, and several lawyers in my family. I've watched as they've seen people in their own profession fall victim to this idea that an LLM will substitute for a paralegal or will write a brief... same sort of atrophy takes place when you become reliant, as opposed to hiring a person who you would train and coach. I think the legal profession is next in the wave of knowledge jobs being devalued, but it's going to be more apparent. Right now when services go down because of bad managerial choices to replace competent humans with LLMs, it's very opaque as to who's responsible.

noncovalence 7 hours ago

I can relate to a lot of this – D&D was never banned for me, but god forbid I ever dare ask my dad for a gaming console. Instead I discovered emulators around age twelve, after what felt like years of saving up for my first Android phone. Naturally, for the the remainder of my childhood, I would spend every single afternoon after school on an IRC server filled with other freaks whose definition of fun was "programming in ARM assembly for 15 hours straight". Sure, it was a hobby in the most literal sense, but by your definition I would never consider classing it as such.

I recently bumped across some of those old friends on GitHub.

The entire game has been decompiled. ROM hacks are written in C now. People are hacking isometric games from 2004 to add 3D with full ray-tracing. They're creating ROM hacks that would've blown my teenage self's mind hard enough to induce a coma. And I consider it a blessing that I got distracted by university applications just in time to never have experienced any of it.

I remember when IDA first released a version that included an experimental decompiler. I remember spending nights on end hunched over my keyboard, cleaning up the output from a single decompiled function just to understand how the game decides when it's going to rain, or when an enemy will attack or defend. Because that was the only way to do it. And now there will never be a reason to need to do it again. I hate that you're right, even assembly programming is just another fucking hobby now.