Hacker News Reader

Comment by LogicFailsMe | Hacker News Reader

Sure sure, but it is surprisingly like managing an intern, or a fresh engineer where there is a pre-existing language barrier. At least in my experience it is. There might be a dash of carefully negotiating with the devil himself in the mix to make sure you ask for precisely what you wish to have built or you get something that meets the spec of what you told it but it isn't what you wanted.

View on HN

And sometimes this intern which can produce code at blazing speeds get some subtleties wrong, so you can never trust their output and have to check every time.

And when you point out the error they sometimes insist what they did is correct, or they confidently "correct" it to something still wrong.

> And sometimes this intern which can produce code at blazing speeds get some subtleties wrong, so you can never trust their output and have to check every time.

You're just describing junior employees lol.

Juniors turn to seniors. You can never trust AI won't glitch...

lol

> Juniors turn to seniors.

AI has gone from "codes like a student" to "fresh hire junior" to "nearly ready for next step in career" at about the same, tiny bit faster, calendar pace as humans.

> You can never trust AI won't glitch...

True.

https://en.wikipedia.org/wiki/Nobel_disease

https://en.wikipedia.org/wiki/Covfefe

https://en.wikipedia.org/wiki/Groupthink

https://en.wikipedia.org/wiki/Boris_Johnson

And?

No, it really hasn't. You still can't remotely trust it to not get stuff wrong.

One of the times I worked with code written by juniors, the entire FAQ was Bob Ross quotes. The rest of the app was in German, and had nothing to do with painting. It also crashed a lot. LLMs, while mediocre, are better than that.

One of the times I worked with someone with a decade or so of experience, their attempt at writing a solution to convert an old file format (loading time: milliseconds) to a database (loading time: milliseconds) would take 20 minutes on some inputs, and he insisted this was the best it could possibly be. In the same day as the standup in which he said this, I'd gotten it down to the milliseconds that was obviously possible.

And there's a reason I linked to those specific Wikipedia articles. There's world leaders whom you can absolutely trust 100% to get stuff wrong on a regular basis, and even Nobel prize winners are not immune to being confidently wrong.

You can't really fully trust humans either. That's why we have processes and checklists; code review; testing at various levels; etc.

With those processes in place maybe we can trust AIs too.

Remotely? Like at all?

Yes, I agree, AI's risk creating a huge pipeline problem.

Babylon 5 Crusade had a device called The Apocalypse Box. It knew everything and you could ask it questions. Unfortunately, it lied strategically so as to both give you an edge here and there whilst it drove you insane with all the lying. I have always seen the similarity to large language models here myself and that was in 1999.

https://babylon5.fandom.com/wiki/Apocalypse_Box

What has surprised me is that JMS has never seemingly noticed this.

thats the reason why I always focus on one function per thread / or one class per thread, and I copy it by hand in VS.NET, to exactly prevent this - never ever I would let it run "unattended" on the codebase :)

To me, it feels closer to dealing with a personal outsourcing agency than working with internal coworkers, because the agents have zero skin in the game.

It's great. You can get a lot of stuff done in parallel. But it's much more a game of checkbox compliance than working with someone who has their head in the same context as you all day. Even a very junior teammate has some situational awareness inside a company/team.

It is exactly that until the moment it isn't and it does something amazing, truly amazing, and I never see it coming when it does happen. But it doesn't happen all that often. It's just that it does happen here and there.

Reply on news.ycombinator.com