Thanks for the article. I've been unwilling to try agentic editing for a while, and this is the kick I need to give it a shot.
My AI skeptic friends are all nuts
861–870 of 1001 posts
Re: My AI skeptic friends are all nuts
#862Fun article, although I'd say I'm maybe slightly more on the skeptical side.
I've found LLMs can't really write a whole program there isn't already a template for. They're good enough at writing individual functions, and as the author mentioned, their breadth of knowledge is better than any individual.
Honestly, I see them as a supercharged version of googling, without all the out of context answers you get from Google since they also place ads and whatnot. I find an LLM can get me an answer far quicker than a regular internet search, even if it's imperfect and I'm probably not going to put it straight into my code.
Re: My AI skeptic friends are all nuts
#863I think this article is pretty spot on — it articulates something I’ve come to appreciate about LLM-assisted coding over the past few months. I started out very sceptical. When Claude Code landed, I got completely seduced — borderline addicted, slot machine-style — by what initially felt like a superpower. Then I actually read the code. It was shockingly bad. I swung back hard to my earlier scepticism, probably even…
I have been using Windsurf for a few months and ChatGPT for a couple of years. I don't feel Windsurf is a massive game changer personally. It is good if you are very tired or working in a new area (also good for exploring UI ideas as the feedback loop is tight), but still not a real game changer over ChatGPT. Waiting around for it to do its thing ("we've encountered at error - no credits used") is boring and flow destroying. Of you know exactly what you are doing the productivity is probably 0.5 vs just typing the code in yourself. Sorry, I'm not going to bang around in Windsurf all day just to help with the training so that "v2" can be better. They should be paying me for this realistically.
Of course, in aggregate AI makes me capable in a far broader set of problem domains. It would be tough to live without it at this stage, but needs to be used for what it is actually good at, not what we hope it will be good at.
Re: My AI skeptic friends are all nuts
#864I think LLMs are great and patient teachers.
But code is something I like to write myself, it's not about speed of producing code. It's keeping it such you still fully understand it.
Re: My AI skeptic friends are all nuts
#865> Meanwhile, software developers spot code fragments seemingly lifted from public repositories on Github and lose their shit. What about the licensing? If you’re a lawyer, I defer. But if you’re a software developer playing this card? Cut me a little slack as I ask you to shove this concern up your ass. No profession has demonstrated more contempt for intellectual property. This kind of guilt-by-association play migh…
That whole section seems so out of place. I don't know why he thinks "The median dev thinks Star Wars and Daft Punk are a public commons" either. I don't know why he thinks the entire software engineering profession is about enabling piracy. I suspect Netflix has more software engineers doing the opposite than every piracy service employs combined.
Re: My AI skeptic friends are all nuts
#866I think this article is pretty spot on — it articulates something I’ve come to appreciate about LLM-assisted coding over the past few months. I started out very sceptical. When Claude Code landed, I got completely seduced — borderline addicted, slot machine-style — by what initially felt like a superpower. Then I actually read the code. It was shockingly bad. I swung back hard to my earlier scepticism, probably even…
For what it's worth: I'm not dismissive of the idea that these things could be ruinous for the interests of the profession. I don't automatically assume that making applications drastically easier to produce is just going to make way for more opportunities. I just don't think the interest of the profession control. The travel agents had interests too!
It's kind of ironic to me that this is so often the example trotted out. Look at the BLS data sheet for job outlook: https://www.bls.gov/ooh/sales/travel-agents.htm#tab-6
> Employment of travel agents is projected to grow 3 percent from 2023 to 2033, about as fast as the average for all occupations.
The last year there is data for claims 68,800 people employed as travel agents in the US. It's not a boom industry by any means, but it doesn't appear they experienced the apocalypse that Hacker News believes they did, either.
I don't know how to easily find historical data, unfortunately. BLS publishes the excel sheets, but pulling out the specific category would have to be done manually as far as I can tell. There's this, I guess: https://www.travelagewest.com/Industry-Insight/Business-Feat...
It appears at least that what happened is, though it may be easier than ever to plan your own travel, there are so many more people traveling these days than in the past that the demand for travel agents hasn't crashed.
Re: My AI skeptic friends are all nuts
#867I’m AI neutral but the writing style here is pretty dismissive, and - to match the tone of the article - annoying as fuck. Most completely reasonable objections to LLMs were totally dismissed.
Seems like very common AI hype post style.
"If you aren't using this AI feature you are left behind."
I usually block people for posting like that.
Re: My AI skeptic friends are all nuts
#868I am finding the most destructive aspect of LLM assistance to be the loss of flow state. Most of the time I can go faster than these tools if I have confidence in myself and allow the momentum to build up over the course of 20-30 minutes. Every time I tab out to an LLM is like a 5 minute penalty over what I could have done unaided on a good day. Getting the model prepared to help you in a realistic domain often takes…
I find the opposite, I can stay in flow state much longer than I could before because tabbing out to an LLM is a tighter loop than me going out to google and searching through sites for smaller tidbits that I have to assemble and get through, rather than a full solution that might not be 100% correct, but most of the time I can get to the working solution quicker.
Re: My AI skeptic friends are all nuts
#869Another day another post about AI by people whose job will be made redundant.
Something fascinating about this HN crowd in particular is its lack of foresight. Companies producing medicine have to take into account longterm effects, yet people working in these “tech” companies push out AI after AI with not a single care in how it’s going to impact people.
Re: My AI skeptic friends are all nuts
#870I think this article is pretty spot on — it articulates something I’ve come to appreciate about LLM-assisted coding over the past few months. I started out very sceptical. When Claude Code landed, I got completely seduced — borderline addicted, slot machine-style — by what initially felt like a superpower. Then I actually read the code. It was shockingly bad. I swung back hard to my earlier scepticism, probably even…
The reason I remain in the "skeptical" camp is because I am experiencing the same thing you are - I keep oscillating between being impressed, then disappointed.
Ultimately the thing that impresses me is that LLMs have replaced google search. The thing that disappoints me is that their code is often convincing but wrong.
Coming from a hard-engineering background, anything that is unreliable is categorized as bad. If you come from the move-fast-break-things world of tech, then your tolerance for mistakes is probably a lot higher.