AI 普及前後:知識工作者對專業價值的反思
Thoughts from Before and After AI
文章作者從個人經驗出發,反思 AI 普及前後,知識工作者對專業價值與人生意義的感受變化。作者認為,AI 將人們累積的知識整理並商品化,令曾受重視的專業技能面臨貶值;以 LLM 取代熟練人手所造成的服務問題,責任歸屬也不容易看清。
Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Thoughts from Before and After AI (chevyclutchfoot.substack.com) 22 points by noduerme23 hours ago | hide | past | favorite | 10 comments help
guitarlimeo54 minutes ago | next[–] 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. reply
rogual1 hour ago | prev | next[–] 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. reply
noduerme1 hour ago | parent | next[–] 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. reply
District55242 hours ago | prev | next[–] 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.) reply
noduerme2 hours ago | parent | next[–] 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. reply
_trampeltier3 hours ago | prev | next[–] 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. reply
noduerme1 hour ago | parent | next[–] 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 ;) reply
herdrick3 hours ago | prev[–] 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. reply
noduerme2 hours ago | parent[–] 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. reply
guitarlimeo34 minutes ago | root | parent[–] 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 reply
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact Search:
來源:Hacker News:AI 熱帖 · news.ycombinator.com