> "Look through the parallel implementations, see which is best (even if it's not good enough), then figure out what changes to your prompt would have helped nudge towards the better solution."
How can non-technical people tell what's "best"? You need to know what you're doing at this point, look for the right pitfalls, inspect everything in detail... this right here is the entire counter-argument for LLMs eliminating SWE jobs...
> How can non-technical people tell what's "best"? You need to know what you're doing at this point, look for the right pitfalls, inspect everything in detail... this right here is the entire counter-argument for LLMs eliminating SWE jobs...
I'm not sure a tool that positions itself as a "programmer co-worker" is aiming to be useful to non-technical people. I've said it before, but I don't think LLMs currently are at the stage where they enable you to do things you have 0 experience in, but rather can help you speed up working through things you are familiar with. I think people who claim LLMs will completely replace jobs are hyping the technology without really understanding it.
For example, I'm a programmer, but never done any firmware flashing with UART before via a USB flasher. Today I managed to do that in 1-2 hours thanks to ChatGPT helping me out understanding how to do it. If I'd do it completely on my own, I'm sure it would have taken me at least the full day to do so, instead of the time it took. I was able to see when it got mislead, and could rewrite/redirect from there on, but someone with 0 programming experience, probably wouldn't have been able to.
It depends on their setup and where they or the LLM gets stuck. If an experienced programmer is there to back them up, then a total beginner could totally make something. That is, given some familiarity with the terminal, specifically the know-how to setup a git repo on GitHub and clone it locally, and then setting up env keys and Aider, and the know-how to run npm I and npm run dev, a non programmer with some terminal skills someone is able to make simple games, purely by talking to Aider using the /voice command. When the LLM or they get stuck is when they'll need some backup from somebody with a decent amount of programming experience to get unstuck. Depending on what their doing though, it's entirely possible they won't get stuck until much further along in the dev process.
I don’t think anyone expects software engineers will disappear and get replaced by janitors trained to proompt. I’m sure experts will stick around until the singularity curve starts looking funny. It’s probably gonna suck to enter the industry from now on, though.
Well, right, how does one become a senior engineer in a world where no one needs to hire a junior? I'm sure many other industries have experiences this already, where the only people who know anything retire and the people are left maintaining a system they could not rebuild such that when something goes wrong the only practicable choice is to replace it with new equipment.
That's where I see AI-written software going, write-once. Some talented engineer gets an AI system to create a whole k8s cluster to run an application and if any changes need to be made, bugs fixed, it will take another talented engineer to come in and have an AI write a replacement and throw out the old one.
Reminds me of this blog, The real value isn’t in the code [0], we're heading for a world that is only code and no one who knows what it does. But maybe it won't matter.
> Well, right, how does one become a senior engineer in a world where no one needs to hire a junior?
You don't. Unless the person is super brilliant I just don't think the industry needs many more new people, there are enough for the next 1-2 decades and after that humans will probably not be needed at all.
People should go where the demand is - medicine, education, policing or whatever it may be.
and on and on, endless thinkpieces about this. Certainly SOMEONE, someone with a lot of money, thinks software engineers are imminently replaceable.
> until the singularity curve starts looking funny.
well there's absolutely no evidence whatsoever that we've made any progress to bringing about Kurzweil's God so I think regardless of what Sam Altman wants you to believe about "general AI" or those thinkpieces, experts are probably okay.
oh I agree but the last three years has felt like an endless chorus of people telling me SWE was going to be obsolete very soon so I had to push back against the idea that "nobody" thinks that.
I wouldn't have to listen to people talk about it all the time if nobody thought it was true
(not GP) To be fair, just because someone says something doesn't mean they believe it. Most of those folks have to know they're being absurd. But I agree saying "nobody" thinks something is over the top. People on the internet can be quite looney tunes.
How can non-technical people tell what's "best"? You need to know what you're doing at this point, look for the right pitfalls, inspect everything in detail... this right here is the entire counter-argument for LLMs eliminating SWE jobs...