Hacker News Reader

Comment by btbuildem | Hacker News Reader

I think the article nails it, on multiple counts. From personal experience, the cognitive overload is sneaky, but real. You do end up taking on more than you can handle, just because your mob of agents can do the minutia of the tasks, doesn't free you from comprehending, evaluating and managing the work. It's intense.

View on HN

For a very small number of people the hard part is writing the code. For most of us, it’s writing the correct code. AI generates lots of code but for 90% of my career writing more code hasn’t helped.

> you do end up taking on more than you can handle, just because your mob of agents can do the minutia of the tasks, doesn’t free you from comprehending, evaluating and managing the work

I’m currently in an EM role and this is my life but with programmers instead of AI agents.

Also EM and it feels like now I have a team of juniors on my personal projects, except they need constant micromanaging in a way I never would for real people.

So you're saying AI doesn't help, and having reports is just like using AI (which you said doesn't help).

What's stopping you from becoming an IC and producing as much as your full team then? What's the point of having reports in this case?

> For most of us, it’s writing the correct code.

I am not sure about this statement, aren't we always cutting the corners to make things ~95% correct at scale to meet deadlines with our staffing/constraints?

Most of us, who doesnt work on Linux kernel, space shuttles, and near realtime OSes, we were writing good enough code to meet business requirements

My point is that coming up with the business requirements was always the hard part (unless you’re writing a scheduler)

Does AI write 100% correct code? No, but under my watch it writes code that is more correct than anything that anyone else on the team contributed in past year or more. Even better when it is wrong I don’t have to spend literal hours arguing with it nor I have to be mindful how what I’m saying affects others feelings so I get to spend more time on actual work. All in all it’s a net positive

I never said it’s not a net positive - I said that writing more code won’t solve the problem.

> under my watch it writes code that is more correct than anything that anyone else on the team contributed in past year or more

This I don’t believe.

I agree.

I provide specific instructions, gotchas when prompting the Agent to write the code. I churn out more instructions quickly by using my voice.

Yes it makes mistakes, but it can correct them quickly as well. This correction loop takes more time if it is a human in my team working.

Started referring to it as "speed of accountability".

A responsible developer will only produce code as fast as they can sign it off.

An irresponsible one will just shit all over the codebase.

This has been my experience too. I feel freed up from the "manual labor" slice of software development and can focus on more interesting design problems and product alignment, which feels like a bit of a drug right now that i'm actually working harder and more hours.

I'm not sure I would agree in totem. Freeing the minutia allows for a higher cognitive load on the bigger picture. I use AI primarily for research gathering, and refining of what I have, which has freed up a lot of time to focus on the bigger issues, and specifically in my case, zeroing in on the diamond in the rough.

> Freeing the minutia allows for a higher cognitive load on the bigger picture

I think we do agree -- the higher "big picture" cognitive load feels more expensive than the minutia cognitive load

do you think this is inherent in AI-related work, or largely due to the current state of the world, where it's changing rapidly and we're struggling to adapt our entire work systems to the shifting landscape, while under intense (and often false) pressures to "disrupt ourselves"? Put another way, if this was similarly true twenty years ago with the rise of Google, is it still true today?

That is fun though.

I hated the old world where some boomer-mentality "senior" dev(s) would take days or weeks to deliver ONE fucking thing, and it would still have bugs and issues.

I like the new world where individuals can move fast and ship, and if there are bugs and issues they can be resolved quickly.

The boomer-mentality and other mids get fired which is awesome, and orgs become way leaner.

Just because there are excess of CS majors and programmers doesn't mean we need to make benches that they can keep warm.

That has more to do with where you work than AI.

Some places have military grade paperwork where mistakes are measured in millions of dollars per min. Others places are 'just push it in fix it later'.

AI is not going to change that. That is a people problem. Not something you can automate away. But you can fire your way out of it.

For sure. I was replying to people not in that, it seems from the commenters here that is where they (and me) have worked or are working now. Whether it is their own company or some other place.

I've only ever worked at places that are at the bleeding edge and even there we had total slackers.

That’s totally orthogonal to what the OP is responding to though. Military software can be bleeding edge as well as extremely susceptible to error prone code which means you need to test more. Similar are the cases with financial softwares which are usually written in ocaml etc. observing your ability to comprehend, the places where you work must be “bleeding” profusely.

> I hated the old world where some boomer-mentality "senior" dev(s) would take days or weeks to deliver ONE fucking thing, and it would still have bugs and issues.

What does that even mean? Are you begrudged manager or enthusiastic youngster who is upset that “boomers” are not killing themselves by juggling thousands of tasks ADHD-style?

"Explain to me like I am five what you just did"

Then "Make a detailed list of changes and reasoning behind it."

Then feed that to another AI and ask: "Does it make sense and why?"

Then get rid of. They can keep 1/10 the humans, and have them run such agents.

Garbage In, Garbage Out. If you're working under the illusion any tool relieves you from the burden of understanding wtf it is you're doing, you aren't using it safely, and you will offload the burden of your lack of care on someone else down the line. Don't. Ever. Do. That.

Reply on news.ycombinator.com