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

AI fatigue is real and nobody talks about it

siddhantkhare.com

Post reply on HN
Print view View original

271–280 of 334 posts

Re: AI fatigue is real and nobody talks about it

#271

Post by johnnyanmac »

zabzonk wrote: post #199
ericmcer wrote: post #63

Earlier quoted context omitted.

Seriously and beyond productivity, flow state was what I liked most about the job. A cup of coffee and noise cancelling headphones and a 2 hour locked in session were when I felt most in love with programming.

Speaking as someone with over 40 years paid programming experience, I've never understood this "flow" thing. I typically do about half an hours typing, get up and walk around, mooch over to colleague and yack bit, or go to the coffee machine, or just think a bit and then go back to the keyboard. Never used headphones - if the environment is too loud, make it quieter. I once moved into a new office area that had a dot…

Half an hour of typing would be above average attention span for the youth these days. That's pretty much how Pomodoro Timers start out for people who can't focus at all.

>if the environment is too loud, make it quieter.

we shifted to open office setups over the decades. There may not even be anyway to make things "quieter" externally.

Re: AI fatigue is real and nobody talks about it

#272

Post by johnnyanmac »

wouldbecouldbe wrote: post #219
well_ackshually wrote: post #54

Earlier quoted context omitted.

This way you can do twice the terrible job twice as fast! (Also, this only applies if what you're working on happens to be easily parallelizable _and_ you're part of the extremely privileged subset of SV software engineers. Try getting two Android Studios/XCodes/Clang builds in parallel without 128GB of RAM, see what happens).

I appreciate sarcasm, but this is just snarky. But yeah improving build speed & parallel running I think are one of the biggest advances devs can do to speed up development time in the AI age. With native apps that can be a challenge. I restructured a react native project to make it faster to iterate, but I have a feeling you might not be fond of rn.

>I appreciate sarcasm, but this is just snarky.

With all due respect, the commment does carry the vibes of "it's okay I can totally multitask" that's nearly a cliche in programming. It just feels a bit insensitive to respond to an article talking about context switching being a big problem with LLMs and then suggest the biggest possible context switch of managing multiple workstations.

And yes, your hardware expenditure will vary per industry. I can barely run one instance of Unreal Engine as is.

Re: AI fatigue is real and nobody talks about it

#273

Post by johnnyanmac »

xnx wrote: post #130
parpfish wrote: post #100

Earlier quoted context omitted.

No, it’s like being a micro manager. I don’t just give somebody a bit’s ticket a let the go. I give them a ticket but have to hover over their shoulder and nitpick their design choices. Tell them “you should use a different name for that new class”, “that function should actually be a method on this other thing”, etc

Nitpicking seems like a choice. It's also possible to be more relaxed/removed and only delve in when there is a problem.

We're at the stage right now where there's always a problem. That's the issue with this type of programming right now.

Re: AI fatigue is real and nobody talks about it

#274

Post by johnnyanmac »

barishnamazov wrote: post #29

This write-up has good ideas but gives me the "AI-generated reading fatigue." Things that can cleanly be expressed in 1-2 sentences are whole paragraphs, often with examples that seem unnecessary or unrealistic. There are also some wrong claims like below: > The Hacker News front page alone is enough to give you whiplash. One day it's "Show HN: Autonomous Research Swarm" and the next it's "Ask HN: How will AI swarms…

>These posts got less than 5 upvotes, they didn't make it to home page. And while overall quality of Show HN might have dropped, HN homepage is still quite sane.

top 24 hours is a better way to get sentiment. Here's the top 5 for today (not including this post at #2)

> DoNotNotify is now Open Source

> I am happier writing code by hand

> Show HN: LocalGPT – A local-first AI assistant in Rust with persistent memory

> Slop Terrifies Me

> Vouch

real shame, we just missed the politics post at #7.

Re: AI fatigue is real and nobody talks about it

#275

Post by keeda »

I have a slightly different theory for AI fatigue. Programming essentially comprises three aspects: problem-solving, coding up the solution, and seeing the solution work. We used to have a balance between them that AI agents are disrupting.

Consider that each aspect has different psychological impacts. The first aspect is the hardest, more intense one.

The second one, translating the ideas in your head into code -- where "flow state" typically comes into play -- is mostly a mechanical, somewhat repetitive task, and hence actually meditative. It is relaxing, almost "cozy", in a counter-intuitive way.

The third part, seeing the code work at the end, is a nice little dopamine reward.

Originally we used to shift between these aspects at a much more balanced cadence: we typically spent a lot of time on the stressful problem-solving part, and even longer on the meditative coding stage. The dopamine hits were spread out across longer stretches of time. (Note that this pattern is fractal, because not only is "high-level algorithm design / architecture -> coding" one cycle, the coding stage itself has similar smaller-scale sub-cycles, e.g. "encountering syntax errors / implementation bugs -> fixing them.")

With LLMs this balance has been disrupted. Now that agents are doing most of the coding, we are spending almost all our time in the stressful problem-solving mode, and then directly jumping to the dopamine hits when the agent delivers code that mostly just works. If you're not YOLOing it and do code reviews, those are also intense, as TFA points out. We get almost none of the meditative coding activity in between.

So it's not surprising that oscillating constantly between two intense modes is going to be much more draining. Vibe-coding may actually be the coding equivalent of doom-scrolling.

I think a lot of recent talk about missing the old way of coding is rooted in the loss of that meditative aspect.

Personally I've noticed I prefer a pair-programming approach with AI, which still lets me spend a lot of my time hands-on coding. It may be because this maintains some of that balance and is more sustainable in the long run. However the allure of productivity from wrangling multiple agents is also very, very seductive...

Re: AI fatigue is real and nobody talks about it

#276

Post by jq-r »

joebates wrote: post #211
alex_c wrote: post #82

Earlier quoted context omitted.

I joke that I'm on the "Claude Code workout plan" now. Standing desk, while it's working I do a couple squats or pushups or just wander around the house to stretch my legs. Much more enjoyable than sitting at my desk, hands on keyboard, all day long. And taking my eyes off the screen also makes it easier to think about the next thing. Moving around does help, but even so, the mental fatigue is real!

I've seriously wondered about merging a home office and home gym into one, and doing sets in between claude working. My usual workout has about 22-30 sets of exercises total and I probably wait on Claude that often in a day. It would be wonderful to be able to spread my exercise throughout the entire day. I'd also include an adjustable height desk so that I could be standing up for much of the workout/workday. I coul…

I have a small gym next to my home office and I just cannot do it. When I train I need to be 100% focused on the exercises otherwise in the best case I'll stagnate, and in the worst I'll injure myself. So instead I do some mindless chores if time allows.

Re: AI fatigue is real and nobody talks about it

#277

Post by johnnyanmac »

xnx wrote: post #151
zkmon wrote: post #58

I said this a few times here. Tech is never to make the life easier for the worker. It is to make the worker more productive and product more competitive. Moving from horses to cars did not give you more free time. Moving from telephone to smartphone did not give more fishing time. You just became more mobile, more productive and more reachable.

How we use efficiency is a choice. It's possible to work a lot less if you accept quality of life from an older era (no phone, Netflix, etc.)

>It's possible to work a lot less if you accept quality of life from an older era

I can throw away all my worldly possessions and post up in a cardboard box. My old bosses would still probably expect me to respond at 8pm to a message. Giving me a company phone if necessary.

It's very well supported that efficiencies have gone up in the workforce over the decades, but hours worked have also increased. We're being squeezed.

Re: AI fatigue is real and nobody talks about it

#280

Post by klibertp »

ramon156 wrote: post #239
stack_framer wrote: post #194

Earlier quoted context omitted.

Wow, this is a new low I did not expect to see. "Just get high and you'll be able to code with an LLM." Preceded by, "I know it's terrible." I'm wholly unwilling to relinquish my health and my morals to "AI" so I can "ship faster." What a pathetic existence that would be.

One of the best programmers I know personally is constantly under the influence of marijuana. As "immature" as it may sound, she's still extremely aware of what she's doing and is able to work in an environment I would give up in after 2 weeks. The kind of environment that denies 1 day PTO for your birthday because of a deadline (hint, every week is a deadline). I do not smoke myself, but it made me realize how littl…

> I do not smoke myself, but it made me realize how little I know regarding THC and CBD

Long-term use causes the psychedelic part of THC effects to diminish over time. At some point, only a mild depressant effect remains - somewhat similar to chamomile. It does have some effect on intelligence and short-term memory, but if the alternative is to be too stressed to think at all, it might be better to just smoke.

Obviously, if possible, psychotherapy or a prescription from a psychiatrist (or better yet, a change of environment) would be better (in the latter case, it depends on the prescribed drug, of course), but THC is not that bad an alternative where it's legal.

Post reply on HN