So compare that plumbing job to say, fixing a bug in code, and you'll be a lot closer. Or go the other way and compare building an app from scratch to plumbing a house from scratch, and again will be a lot closer.
Plumbers in my parts charge about hourly what a dev makes. Making an app just takes a lot longer than whatever plumbing job is being referenced.
If I'm coming to fix a sink, I have a pretty good idea about how it will have been done. That's not the case for fixing a bug in an unseen codebase.
This comparison makes no sense.
I don't understand why this guy is the CEO of OpenAI.
Imagine instead, three years ago, before ChatGPT and Dall-E, he tweeted something like:
“”” 2019: $30,000 to (produce real game assets out of my programmer art sketches)|(write a well researched biography)|(animate a short film)|(narrate my documentary), $300 for a plumbing job.
i wonder what those relative prices will look like in 2028! “””
We’re not there yet, but already it’s pretty clear the prices of the former will decrease dramatically, to the point it takes a similar amount of human effort, training, skill, and experience as the plumbing job - then perhaps get even easier!
It is interesting to think why most other comments try to pick apart the original tweet instead
Building a iPhone app will become relatively cheaper (in context of current trends like ChatGPT and Copilot contributing to drastic improvements in developer productivity)
Physical labor and skills (that today's tech is not on the verge of disrupting) will remain in demand and become relatively more expensive to procure.
So he seems to be predicting a kind of reversal of these data points in near future.
Personally I agree with other comments here that say these two are not comparable. A $300 plumbing job today is probably more like a bug fix in software terms (a bug fix being much cheaper than $30K , may even be less than $300 already) rather than a full app build.
Price of training GPT? 30k$