Yet?

this is so true. but watching how much better the car driving algos have gotten, and midjourney, i do think that it is possible to train them for taste. i think this is the missing piece/trick that will take another 5-10y to eek more excellence out of them tbh

This presumes that LLMs can't reason, which appears to be untrue

Vibe coding is like Canva vs illustrator. You’re gonna have the exceptional people that use the latter well but most people just need canva to get their point across.

Mona Lisa probably had a sketch or two ¯\_(ツ)_/¯

🎯

I’m not a programmer by trade, but I agree with your points. Quality is important especially for the longevity of a project.

I think vibe coding has more of a temporary/prototyping use case. Think of it as something similar to a disposable camera or a 1 time event use case. I use it to bring my creative ideas to fruition, to see if one is worth it enough to stick to actually put forth long term effort into it. It reminds me of a blank canvas.

Also Damus is great.

if you can't vibe with codes, I'm glad I'm a musician instead lol

Is this a forever statement?

Interesting point about engineering excellence! Similar to how you'd want flawless systems for critical operations, our ​​buy yacht​​ options at https://www.yachttrading.com/ represent the opposite of 'average' - pure craftsmanship on water."

Maybe AI is more than the sum of its parts.

I would love to hear what ffmpeg guys think about vibe coding.

Vibe-coding reduces the cost and time of prototyping by at least a factor of 100. I expect that two good developers can do the work of a team of 20 in dev, test and prod, in a tenth of the time. AI is a productivity tool, like moving from calculators to spreadsheets, just 100x better.

That's not true. It would be like saying that some human reading tons of books will only have “average” writing capabilities.

Learning is an additive process. Specially with complex neural networks and the training algorithms used. When you learn something you don't necessarily need to unlearn something else to make room for it.

It's like LLMs having some kind of intelligence cap. But given that cap, they max it out and perform way better than the equivalent human.

That being said, most vibe coding quality sucks. But not because of the reasons you mention.