Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI;DR (AI; Didn't Read)

rickmanelius.com

Post reply on HN
Print view View original

651–660 of 729 posts

Re: AI;DR (AI; Didn't Read)

#651

Post by aizk »

aaaronic wrote: post #252
LPisGood wrote: post #56

My coworkers continue to dump hundreds of lines of AI documentation in every PR and every other line of code has between one and ten lines of AI generated comments, talking about the real unlock and how things are byte for byte identical on the load bearing path or how the acceptance ladder is misleading. Features are coming out and metrics are improving, but we’re basically in a post readability code base, with the…

My "favorite" Claudism is when I critique its work and ask it to remove some unnecessary part of the design -- and then the diff has more green than red because it added comments about why the code is no longer there -- the code that was never in the mainline and never asked for!

Opus 5 was so bad at this I switched back to Opus 4.8

Re: AI;DR (AI; Didn't Read)

#652

Post by BeetleB »

watwut wrote: post #576
BeetleB wrote: post #221

Earlier quoted context omitted.

> No, linking blog post and posting AI generated content is not the same. One is ok, other is not. Meet "the masses". Sorry, but no. Sending me a link as an alternative to having a conversation is totally not OK.

I recommend this great book, blog or whatever is completely fine in the middle of the conversation. And it is not the same as "I had AI generate this".

> I recommend this great book, blog or whatever is completely fine in the middle of the conversation.

Please reread my comments. That's not the scenario I was discussing.

Re: AI;DR (AI; Didn't Read)

#653

Post by corndoge »

a2ff6eeb0 wrote: post #467
delecti wrote: post #170

Earlier quoted context omitted.

It's not the ocean, it's the poster's own team. A simple "AI comments suck" in a sprint retro would be trivially easy and would at least start the conversation.

It's the management desired direction, if the place is anything like my workplace. Everyone is all in on AI, and if you're not using it for everything possible, you're on the chopping block. And, really, it works. You can copy and paste between tickets and Claude, and then do manual testing. Then you tell Claude to self-review for clarity and minimalism, and stop worrying. Sure, today it's not as good as a human, but…

Seems like a sad way to spend the limited hours of your life

Re: AI;DR (AI; Didn't Read)

#654

Post by BeetleB »

27183 wrote: post #440
LPisGood wrote: post #327

Earlier quoted context omitted.

What I’m trying to say is that they can generate many more hundreds of lines of code and comments than I can reasonably review. Maybe the comments do make sense they’re just five lines longer than they ought to be. Multiply that by 20 times and then add in 150 lines of documentation that is not technically incorrect. Not to mention that the response to each review will come with its own set of new comments and new do…

Your bosses don't want you actually reviewing it anyway. They want you to approve the PR . Just give them what they want. I'm currently not working in tech, but I will again sometime within the next year or so, and I've been reflecting on my career in light of the recent AI madness. I think the biggest mistakes I've made over the last decade+ in tech as a software engineer have been caring . I've worked at small, med…

> Your bosses don't want you actually reviewing it anyway. They want you to approve the PR. Just give them what they want.

Then remove the whole process and let it go through automatically. I've worked in high trust teams where we would do that - there was no gate.

Re: AI;DR (AI; Didn't Read)

#655

Post by BeetleB »

eudamoniac wrote: post #414
BeetleB wrote: post #409

Earlier quoted context omitted.

The bottleneck existed before LLMs. Amdahl's Law applies. Just because code can be produced 10x faster doesn't mean much if other parts of the flow can't be sped up. As I said here and elsewhere: If management is concerned code review is a bottleneck, let management be aware that the process has to change. Either the human is not as thorough, or he uses some combination of his brain + LLM to conduct the review. It's…

Management is not a logical formula. Management says all code must be reviewed, your coworkers are approving (rubber stamping) PRs 10x faster than you, why aren't you able to keep up? Why are you being difficult?

Again: I'm failing to see how this is different from pre-LLM days. If you have coworkers who are sloppy in reviewing, then you would have been a bottleneck compared to them back then as well.

Do you not have team meetings with the manager? Examine some of the PRs your coworkers reviewed, find the flaws, bring it up to the whole team, and get buy-in from the manager that he's OK with it. If he is, stop worrying and let the PRs through.

Also, start writing those CYA emails.

Re: AI;DR (AI; Didn't Read)

#656

Post by 27183 »

BeetleB wrote: post #654
27183 wrote: post #440

Earlier quoted context omitted.

Your bosses don't want you actually reviewing it anyway. They want you to approve the PR . Just give them what they want. I'm currently not working in tech, but I will again sometime within the next year or so, and I've been reflecting on my career in light of the recent AI madness. I think the biggest mistakes I've made over the last decade+ in tech as a software engineer have been caring . I've worked at small, med…

> Your bosses don't want you actually reviewing it anyway. They want you to approve the PR. Just give them what they want. Then remove the whole process and let it go through automatically. I've worked in high trust teams where we would do that - there was no gate.

If that's what they wanted they'd mandate that process. They want the deniability of having you rubber stamp the PR. That way they can blame you when it goes wrong. There may be high trust teams, but I've never seen one firsthand, but my sample size is only N=5.

Re: AI;DR (AI; Didn't Read)

#657

Post by keeda »

0x69420 wrote: post #474
keeda wrote: post #473

I dunno, not a huge fan of TL;DR (even though it definitely has its place for long-winded walls of text) -- I feel like TL;DR is a cause or a symptom or both of our dwindling attention spans. I think this inability to read anything longer than a tweet, and the resulting loss of nuance, has been pretty bad for discourse, online or off. AI-writing is definitely very annoying to read, but I don't know if it's ideal to s…

but our dwindling attention spans are caused by a deluge of material sapping our energy before we can devote it to something worthwhile. developing a fast and frugal mental filter for insubstantial walls of slop is a new prerequisite for not being so exhausted that you default to saying tl;dr to the real, good stuff.

Kind of yeah, but I think more than just the deluge of material it is the nature of that material: even before AI slop, social media has inundated us with short-form content via their algorithmic firehose. Our attention span is dwindling because we have been trained to mentally switch off if we don't get a hook or a dopamine hit within the first few seconds, at which point we continue doomscrolling. (Indeed, studies suggest that attention spans in both kids and adults have reduced drastically in the last decade coinciding with the rise of social media.)

A quick mental filter is absolutely needed though. But we should consider tuning that filter for signals other than AI writing. I've built one up where a quick skim through the post tells me if there is any substance there or just some shallow "it's not A, it's B" punchline. It took some work to tune that filter, but has let me find some very interesting articles that I otherwise might not have read.

Re: AI;DR (AI; Didn't Read)

#658

Post by queenkjuul »

LPisGood wrote: post #56

My coworkers continue to dump hundreds of lines of AI documentation in every PR and every other line of code has between one and ten lines of AI generated comments, talking about the real unlock and how things are byte for byte identical on the load bearing path or how the acceptance ladder is misleading. Features are coming out and metrics are improving, but we’re basically in a post readability code base, with the…

Personally I'm criticizing every single comment until I quit the industry or my coworkers convince their Claudes to shut up.

Maybe I'll sneak in a change to CLAUDE.md telling it to stop

But half the time these paragraph-long comments don't even make sense, or refer to a previous iteration of the branch that has since been deleted and nobody including the author ever looked at

Re: AI;DR (AI; Didn't Read)

#659

Post by queenkjuul »

koe123 wrote: post #516
koyote wrote: post #263

Earlier quoted context omitted.

It's not just claude, all AI is unable to produce something concise. On the surface everything looks 'good' whether code or prose, but then if you dig a bit, try and understand the whole text you quickly realise that 80% of it is unecessary and the whole thing could have been re-worded/re-coded into something a fraction of its size and complexity. I asked Sol to reduce the length of some documentation we had by makin…

I have a theory that AI code looks good because you never subsequently come up with your own alternative.

I had Claude knock out a feature but it was too big of a change set for a single PR so i set up a worktree and started extracting parts piecemeal. Along the way I would clean up and rewrite pieces. By the end of this process i had abandoned everything Claude wrote and looking back at the original branch I was like "oh my god, that's so bad in so many ways, i can't believe i was actually just gonna ship that"
Post reply on HN