Crumb
c/fuck_ai · by vividspecter@aussie.zone · 25d

A Scientist Was Presenting His Vibe-Coded New Work When One of His Colleagues Pointed Out Something Extremely Embarrassing

https://futurism.com/artificial-intelligence/scientist-horrified-vibe-coded-work-conference-wrong

102 points · 17 comments · view on lemmy.world

17 Comments

churrodestroyer@lemmy.zip · 102 pts · 25d (3 replies)

But just ten minutes into his presentation, a collaborator interjected telling that “something seemed off to them.” As it turns out, the way the new algorithm handled the edges of the surveys was “wrong,” Sutter admitted.

“It wasn’t a typo, and it wasn’t a missing citation or a factor of two,” he wrote. “It was subtle, but it was very wrong, and everything downstream of it was also wrong, and I had shared the whole thing in a room full of people who trusted me.”

"It was very wrong" HOW?!?!

The article might as well have continued "It was so obviously wrong that to go into detail would waste our readers time as any reader of this publication is well aware of the kinds of things that go wrong with the edges of surveys."

jesusactuallyhateshorses@sh.itjust.works · 55 pts · 25d

I've found most "articles" from futurism to be completely devoid of detail. Just blurbs and clickbait.

spicehoarder@lemmy.zip · 12 pts · 24d (1 reply)

Fucking parallelisms. Clanker wrote this article.

november@piefed.blahaj.zone · 5 pts · 24d

The guy himself at least used an LLM for his quotes.

leadore@lemmy.world · 60 pts · 25d (12 replies)

If you read the scientist's original article (linked to from this article), he seems to have learned very little from the experience, except to be more careful while continuing to use the LLM. He compares it to alchemy, which he dubs "pre-science".

He says:

But the story doesn’t end with my humiliation. Alchemy was already on my mind, and the more I read about it, the more I realized the alchemists’ work wasn’t an embarrassment to science, it was the road that led to it. They never found the philosopher’s stone. But they found something else, something we need really badly right now.

(though he doesn't tell us what the 'something else' was). Then later,

But dang it, LLMs are so undeniably useful. I have been writing code since I was 5 years old. I haven’t written a line of it since January. And for 20 years I did my own research, which is to say I read the papers, chased the citations down their rabbit holes, and built every argument myself. AI agents do all that now.

These were not lazy trades. They were correct trades, and I’d make them again tomorrow. Somewhere in the past year an LLM became my thinking partner, and I say that without embarrassment. I’m not the only one. Hundreds of millions of people reach for these tools every day, and they are not idiots.

soratoyuki@piefed.zip · 58 pts · 25d (1 reply)

Somewhere in the past year an LLM became my thinking partner, and I say that without embarrassment.

You're good, I have enough secondhand embarrassment for the both of us.

stoy@lemmy.zip · 15 pts · 25d

Somewhere in the past year an LLM became my thinking partner,

That statement terrifies me, I very seldom use LLMs, maybe one question in a month, and that question is usually when I have exhausted the other alternatives.

Example, last week I attended an airshow and needed to program my radio scanner to a local radiostation where they rebroadcast the announcer to enable you to hear it clearly rather than through the annoying speakers.

And I just could not get my Icom IC-R6 to work, it would randomly change radio modes as I was tuning, I also had trouble setting the tuning steps correctly.

I read the manual, but I could not figure it out, so I caved and asked chatgpt, and got the help I needed in quick order.

I am certain that I would have figured it out on my own, but it would have taken far longer.

Anyway, I am always careful to limit my interaction with LLMs/AI as I start feeling my skills stagnating and even reducing as I use it.

Witchfire@lemmy.world · 29 pts · 25d (1 reply)

So he's been lost to AI psychosis. Cool.

leadore@lemmy.world · 4 pts · 24d

I'm afraid so. The overall tone of his article is a bit manic and he seems really excited about his "alchemy" metaphor, as if its history somehow predicts that AI will become a good, useful thing--while repeatedly admitting that alchemy was totally unrealistic and a complete failure.

frustrated_phagocytosis@fedia.io · 25 pts · 25d (2 replies)

"Hundreds of millions of people reach for these tools every day, and they are not idiots" Citation needed

DelnitaCrane@lemmy.world · 15 pts · 24d

Hundreds of millions of people reach for cigarettes every day, and they can't be wrong

Tar_alcaran@sh.itjust.works · 7 pts · 24d

Yeah, have they met people? We're absolute morons.

gedfromgont@piefed.ca · 24 pts · 25d

It really makes me sad to see actual people who are not under corporate pressure, write this out.

x00z@lemmy.world · 17 pts · 24d

Somewhere in the past year an LLM became my thinking partner, and I say that without embarrassment. I’m not the only one. Hundreds of millions of people reach for these tools every day, and they are not idiots.

Fucking idiot.

Arola@sh.itjust.works · 16 pts · 25d

The last couple of lines has me thinking this is satirical/sarcasm:

"...Following his embarrassing slip-up in February, Sutter vowed that he works “differently now” and is ready to immediately “distrust” anything an AI says" and "...when we pasted his latest piece for Nautilus into AI detecting tool Pangram — which is far from perfect — it informed us that 56 percent of the text appeared to be written by an AI".

SnoopSqueak@lemmy.today · 12 pts · 24d

AI agents do all that now.

But do they do it correctly? Who is checking?

These were not lazy trades.

Saying this immediately after "I used to do all the work myself, but now I don't."

They were correct trades, and I'd make them again tomorrow.

"I'd rather be wrong than do my own research."

Somewhere in the past year an LLM became my thinking partner, and I say that without embarrassment.

Does he know LLMs don't think?

I’m not the only one. Hundreds of millions of people reach for these tools every day, and they are not idiots.

Did AI tell him that? Idiot.

Appealing to popularity suggests he is not thinking for himself. Too many people are outsourcing their thinking to unreliable software in much the same way that religious people outsource their morality to unreliable books.

HeyThisIsntTheYMCA@lemmy.world · 9 pts · 24d

Wow. He fucked up, in front of his colleagues, and still drew the wrong conclusion?