> Any time I notice something in code review and have to get the agent to fix it.. I throw it on the list! My list is like 200 items now.
This is a gripe I've had with AI tools for a while now. Though it's gotten somewhat better in time, but we don't really know what to expect from the tool in terms of quality. Ex. I'd expect a human engineer to probably not use a brand new assertion library for a new test when there are 200 tests using an existing one. But Claude has done this to me multiple times. So I have to add yet another item to the list, like you have, and tell it to look for testing conventions before writing. But, there is plenty we don't have to tell it, like what a function is or a test should probably cover the change in the diff. But we don't really have a list of what things are on each side so we're just left to sort of hunt and peck to build a viable solution.
I've had this exact sentiment in the past couple months after seeing a few PRs that were definitely the wrong solution to a problem. One was implementing it's own parsing functions to which well established solutions like JSON or others likely existed. I think any non-llm programmer could have thought this up but then immediately decide to look elsewhere, their human emotions would have hit and said "that's way too much (likely redundant) work, there must be a better way". But the LLM has no emotion, it isn't lazy and that can be a problem because it makes it a lot easier to do the wrong thing.
>The engineers who thrive will be the ones who can resist the temptation to over-engineer when the marginal cost of adding complexity drops to near zero.
I think this isn't being discussed enough in the SWE world. It wasn't too long ago that engineers on HN would describe a line of code as "not an asset but a liability". Now that code is "free" though, I'm seeing more excessively verbose PRs at work. I'm trying to call it out and rein it in a bit but until engineers on average believe there is inherent risk here, the behavior will continue.
I don't see how the rich don't understand this. Guess they just think it won't happen to them and they'll be smart enough to cash out of America and go elsewhere before everyone else looses.
There's an older article that gets reposted to HN occasionally, titled something like "I hate almost all software". I'm probably more cynical than the average tech user and I relate strongly to the sentiment. So so much software is inexcusably bad from a UX perspective. So I have to ask, if code will really become this dirt cheap unlimited commodity, will we actually have good software?
> To be fair, I was able to get it to work pretty well after giving it extremely detailed instructions and monitoring the "thinking" output and stopping it when I see something wrong there to correct it, but at that point I felt silly for spending all that effort just driving the bot instead of doing it myself.
This is the challenge I also face, it's not always obvious when a change I want will be properly understood by the LLM. Sometimes it one shots it, then others I go back and forth until I could have just done it myself. If we have to get super detailed in our descriptions, at what point are we just writing in some ad-hoc "programming language" that then transpiles to the actual program?
If there's one thing both parties agree with, it's that you can't ever vote for a third party because that's effectively voting for the other major candidate. So the problem of not having more than 2 choices perpetuates indefinitely.
> AI means that those 'easy' tasks can be automated away, so there's less immediate value in hiring a new grad.
Not disagreeing that this is happening in the industry but it still feels like a missed opportunity to not hire juniors. Not only do you have the upcoming skill gap as you mention, but someone needs to instruct AI to do these menial/easy tasks. Perhaps it's only my opinion but I think it would be prudent to instead see this as just having junior engineers who can get more menial tasks done, instead of expecting to add it to the senior dev workflow at zero cost to output.
If you were a dev selling a game years ago when physical distribution was the only method, you'd likely end up with a lot less than 70% after both the publisher and retailer take their cut.
I don't have access to the actual study but I wonder if they had access to any metrics for the individuals that could quantify the intensity of the insomnia. Chronic insomnia/sleep deprivation is already correlated with heart disease. Could it be that the people with worse insomnia were more likely to take melatonin but also more likely to have heart issues due to the insomnia alone?
I feel like hosting in NYC is even more of a public service given that space is limited and not everyone has a living situation suited for it. Props to you for making it happen. Been doing what I can here as well. Cheers!
This is a gripe I've had with AI tools for a while now. Though it's gotten somewhat better in time, but we don't really know what to expect from the tool in terms of quality. Ex. I'd expect a human engineer to probably not use a brand new assertion library for a new test when there are 200 tests using an existing one. But Claude has done this to me multiple times. So I have to add yet another item to the list, like you have, and tell it to look for testing conventions before writing. But, there is plenty we don't have to tell it, like what a function is or a test should probably cover the change in the diff. But we don't really have a list of what things are on each side so we're just left to sort of hunt and peck to build a viable solution.