Hacker News new | past | comments | ask | show | jobs | submit
I love this comment.

SWEs who oppose AI seem to forget that they are paid to solve problems, not write code.

Most SWE who "oppose AI" in my experience, don't oppose AI but oppose its how its being adapted or used. All gas no brakes. You can't really say "lets think about it a bit" without becoming AI denier/opposer or whatever.

We make aggressive decisions, based on overly optimistic bet on "AI will fix it" mindset. I think it should be more towards, given what we have now and may put in little bit of optimism around AI how can go forward without completely dismissing decades of engineering principles we developed. If AI really goes 10x better than what it is now, then we can just build stuff much faster then anyway, but it didn't we made something reliable today as well something we can keep building on 5/10 years from now.

> You can't really say "lets think about it a bit" without becoming AI denier/opposer or whatever.

What does "lets think about it a bit" even mean though? You want other people to stop using productive tools? Or you want to be able to opt out of using productive tools without an employer worrying about it? There always seems to be either a shifting of burden to others, or a denial that AI tools increase productivity.

People can think about AI and use AI with all gas and no brakes. Or you can opt out of using AI for as long as you like to think about things.

loading story #50569060
loading story #50568784
loading story #50568894
loading story #50568988
loading story #50570898
Everyone is paid to solve problems. If the things that make you stand out at solving problems compared to the average worker is writing code, this no longer being a moat is probably not good for your bargaining power.
The things that make me stand out at solving problems is that I can effectively convince machines to solve problems at massive scales.
loading story #50568360
I’d wager a lot of people not from software engineering will be just as good or better than software engineers for that, if only because most problems in business are business problems: Joe from accounting with a solid understanding of accounting and passable prompting might be better at solving accounting problems with AI than John the dev with limited accounting understanding but great effectiveness at pointing the AI towards wrong problems.
loading story #50570464
loading story #50568889
loading story #50568395
I actually think a wide majority of SWEs, pre-AI, were in fact primarily paid to write code.

Most people don't do this for the love of the game.

Large corporations needed java code as a means to an end and an army of workers needed a paycheck. That dynamic is coming to an end.

This. The JIRA mills where folks log in, say "no blockers" in a half-hour standup, fix an endless stream of tickets, then log off and go do what they actually want to do. That's over, thankfully.
Many SWEs will not be paid to solve problems anymore due to AI. The ones that remain might be better equipped for the task, although I don't expect interests to align on doing a better job instead of a cheaper one.
There's room for nuance there. I am well aware that the goal is to solve problems, but I actively enjoy the process of designing a program and writing the code.

Sure, maybe I can more effectively and efficiently solve problems by driving an LLM instead, but that's not a job I enjoy, or even care to do professionally.

System design has always been the most important and interesting part of software development. Now you can do that at a scale never seen before, it’s awesome.

That being said, software (written by hand) has replaced many craftsmen. It’s a bit hypocritical to bemoan AI taking away our need to write code line by line. It’s what we’ve been doing to others since the first computer started computing.

loading story #50569098
loading story #50568499
That can be interesting but LLMs can do system design fine. Sure it can make crap but it doesn't take 10 years of experience to learn the 50 or some prompt variations to get it to sort things out. As time goes on those will be baked into the harness.

People on hacker news like to think they are Picasso but the reality is 50% as good in 5% of the time has always been better from the business perspective.

SWEs also often forget that most of the software they create was designed to take jobs away from other people in other departments.
The frame often is positioned to include yourself and your own interests at the center of it.
in my 30ish years of existence I haven't seen software take away people's jobs. transform yeah but take way jobs ?

unless when combined with robots.

When a pawn on a chessboard takes a piece is do you blame the pawn or the chess player?

Software devs were guilty of wanting one of the few middle class jobs left. Nobody enters the profession cackling evilly about how many jobs they will take.

loading story #50568045
loading story #50568250
I've always said that software developers are the perfect middle class for capital: they never fight for their own interests and find it fun to put down their peers as a pastime. Even then, capital came gunning for them and then what did they do? Beg "HARDER, I'M NOT LIKE THE OTHER WORKERS". It's incredible really.
> SWEs also often forget that most of the software they create was designed to take jobs away from other people in other departments.

Often maybe, but not always.

Avoiding displacing human labor that doesn't want to be displaced is always something I've thought about before any job I've taken.

No putting clankers before humans, no grifts, scams, pushing ads, gambling or incentivizing other forms of addictive behavior.

Fairly simple rules to live by, though it does mean there are entire major companies (and entire market categories of companies) I could never work for in good conscience.

I do find the fact that AI is disrupting us software developers first and foremost seems karmically fitting in general.

I'm an swe who opposes AI. I don't think we should be spending so much non-renewable sources of energy on AI. These energy sources are taking from our future.
I'm very pro-AI but I also think we should be using this opportunity for energy transition.

Some countries are making laws about renewables requirements for datacenters. Generally they are structured as "gas is allowed as a backup but there are big fines if it is used more than X days a year", and then let the market decide what mix of solar + wind + batteries wins.

The US should do this too.

A developer's energy use from AI is generally tiny compared to their energy use from driving or eating meat.
loading story #50568035
loading story #50567626
loading story #50567433
The energy used by AI is, frankly, negligible. The water too. It’s a rounding error as compared to agriculture and manufacturing. The water use of all AI data centers combined doesn’t even reach the change in water usage over the past year for agriculture. Never mind the base usage, which is orders of magnitude greater.

Moreover, big picture, software is getting built anyway. It’s not like if you don’t use AI then that energy use is gone. You need devs no matter what.

loading story #50566838
Success at developing AI will save more energy than you can imagine. Half the cars on the road won't need to be there, if nothing else.

Edit: never mind what you don't want to see happen: what do you want to see happen?

loading story #50570210
loading story #50569427
Most of the SWEs I see opposing AI, are opposing it because it doesn't solve the problem. It creates patches that introduce more problems, faster, than if they just did it themselves.
I’m paid to be a meat proxy because our problems are trivial. The only thing i offer is writing the prompts that say “implement this” and “you overcomplicated x, make it simpler” occasionally including the way I want it fixed. Important steps in the process and skill and experience still help not pushing slop but I’m not really “solving problems” anymore.
You are solving the problem of someone else having to do that exact task.

That is why I think AI is more of a service than a tool.

I am not sure what real problem my barber is solving. I am physically capable of using scissors and clippers and learning how to cut my own hair. But at the end of the day, I rather just pay someone trustworthy to do it instead.

The money I pay is worth it for me to not have to deal with any of the previously mentioned tasks.

loading story #50569026
loading story #50570329
SWEs aren't paid to solve problems (most of the time). They are paid because they are a necessary evil sitting between product owners and revenue. It doesn't actually matter that much if problems are solved. Their goal is to increase revenue, and there is not a linear correlation between solving problems and making revenue. Business leaders would absolutely love it if they could take a dump and have enough people buy their shit and eat it while not having to pay for the SWEs that they see as middle-men. If SWEs were paid merely to solve problems, the world would be a much different place. Corporate software engineering in the 2020s is as much about solving problems as advertising is about communicating the merits of a product.
Every single person, every product, every single "cost centre", everything is the necessary evil between "owners" and revenue. We should just optimize it all away and let "owners" print their money to infinity since that is the only thing they value and there is no need for anyone else to stand in their way by reducing their revenue. Of course, their money will be worthless but why is that a problem, they have it all! On the other hand, if you take the view that "economy" and therefore companies and products exist to help people then value != money. SWEs as they exist now already belong to 2 groups - "code shovellers" and problem solvers - same as any profession. LLMs can be used to shovel more code or help solve problems, most people will just shovel code (just look at python or javascript ecosystems) and few will continue to solve problems. Once LLMs can shovel code well enough then they will replace all other "code shovellers" but LLMs will never be able to solve problems, at least not on the same level as people. But I am sure that will not stop "owners" from thinking they can so every profession will be obsolete. Then there are two alternatives - everyone becomes a slave or we enter "Star Trek" style universe. I know which alternative is more likely (remember "Star Trek" is fiction).
loading story #50569244
I don't mind solving problems with AI. I'm worried about getting fired because they think they can do the job without me entirely.
SWEs where paid to solve problems by writing code.

Solving problems is not distinctive characteristic of a software engineer, doing it by writing code was.

"Solving problems" is a meaningless thing to say, and seems to be the next evolution of "full-stack engineer". It's a result of dissolving what used to be a specialty with its own subspecialties. Make the role of an engineer as generic as possible so they become interchangeable, easier to replace, and eliminated when possible.

Plenty of software engineers are paid to not solve problems. Some are, but I would say that most are paid to do the thing. Orgs beyond the startup phase don't actually like problem solving because it introduces risk of instability. Just doing the thing isn't the same thing as solving problems. Business leaders would be very happy if they could eliminate the human from the doing the thing step, and not all of us can simply become the problem solvers because that's infeasible.

We certainly solved the problem of companies having to hire as many of us or pay us as well.
Some people resent that the problems they used to solve were technical, but overnight, turned into managerial.
People who write comments like this forget that the product AI is selling is automated problem solving. In fact, generalized superhuman problem solving is the only way to justify their valuation.

Your comment is stating a belief that Anthropic will fail as a business.

If only. AI pilled SWEs seem to have not noticed that the only problem "building agents nobody needs" is solving is how to satisfy an investor with FOMO.

I never thought I'd miss the days when executives used to rabbit on about customer obsession but the new fashion of AI psychosis is worse in just about every way.

I hate this comment.

AI-pilled folks seem to forget that SWEs are paid to develop _software_, which last time I checked was made from code.

No one merely "solves problems" in a vacuum. You have to engage with a domain, a medium, and the tools of both.