AI is removing the middle class of software engineering? • Hacker News Board

Re: AI is removing the middle class of software engineering?

#921

Post by roenxi » Thu, Aug 13, 2026, 9:37 PM UTC

breezybottom wrote: post #871

roenxi wrote: post #819

Earlier quoted context omitted.

You can also buy from companies that hire licensed software engineers if you want to? The main thing that would stop people patronising licensed individuals is if there was no obvious skill gap between a licensed and unlicensed practitioner that justifies the price gap. People don't need a government body to tell them to check for credentials. If people care they can ask (I had to give a copy of my degree certificate…

That's the absurd libertarian logic. I'm supposed to be able to assess the credibility of whatever credentials I'm shown? Despite having no expertise in the field?

The government can sponsor a credential if that helps you? The libertarian position isn't that the government can't have a position on what the standard should be, just that people who think there is a more cost-effective standard should be allowed to use them.

Much like how you can hire a programmer based on that you think their work is good or they made a lot of money programming, rather than just limiting yourself to the people who went to some fancy university. If you don't feel competent at assessing programmer quality then you can exclusively hire ex-Google or MIT programmers or what have you and pay an absurd premium for the most credentialed US developers. But most people don't need to do that.

Re: AI is removing the middle class of software engineering?

#922

Post by codexon » Thu, Aug 13, 2026, 10:23 PM UTC

Vegenoid wrote: post #911

codexon wrote: post #661

Earlier quoted context omitted.

The output is indistinguishable from manually written code.

"My vibecoded stuff is indistinguishable from handwritten stuff" and "I got anti-AI comments after vibecoding my website even though I didn't say anything about AI" are not very compatible with each other.

There was absolutely nothing showing that it was vibe coded other than the fact that a website that didn't get any updates for 5 years suddenly got a 20+ bullet point update. People just assumed it was vibed because of the sheer number of changes.

There was not 1 mention about anything being broken, they were simply complaining about AI.

Anti-AI psychosis is a very real thing that is happening https://x.com/Jediwolf/status/2054776716770320631

Re: AI is removing the middle class of software engineering?

#923

Post by WalterBright » Thu, Aug 13, 2026, 10:31 PM UTC

rightbyte wrote: post #782

WalterBright wrote: post #747

Earlier quoted context omitted.

True, but it shows a willingness and ability to do the work.

Such recruiting selects for KPI hunters. It is probably bad for an org to select for people that game it rather than doing "what is right". A dice throw and some chitchat would be better recruiting.

These arguments all sound familiar to the "high stakes testing" debate about the SATs. The tests were unfair, did not measure college readiness, could be gamed, etc.

Then universities dropped the SAT requirements. Even MIT was suckered into dropping them. It was a disaster. It turns out the SATs were a very strong predictor of college success. MIT and others reinstated the SATs.

Re: AI is removing the middle class of software engineering?

#924

Post by WalterBright » Thu, Aug 13, 2026, 10:38 PM UTC

tombert wrote: post #763

WalterBright wrote: post #727

Earlier quoted context omitted.

You might be surprised at how incompetent some engineers are. They get filtered out with the leetcode test. Of course, it doesn't filter them all out, and that's why there's a followup interview round.

I have given dozens of interviews at big companies and small companies and medium companies so I am not speaking out of my ass. These leetcode tests are not good filters, nothing is “surprising” me. Generally they are just wanky bullshit from some middle manager’s slightly-incorrect recollection of their first “data structures and algorithms” class, and the solution is almost invariably “use a hashmap” or “use a minh…

[dead]