On one hand you’re right, but you’re missing the other side - business requirements have always expanded with each innovation.
Basic web dev has been commoditized for awhile, but a company of any size requires much more than basic web dev.
Humanity has an uncanny ability to devour any excess capacity, and I see no reason it won’t do the same with things AI makes easier/lower cost.
What I do worry about is that AI will put further pressure on the lower skilled jobs keeping many people afloat. Call centers for example, likely will not exist very soon.
"Humanity has an uncanny ability to devour any excess capacity"
Well put!
You could also see the pressure on lower skilled jobs as a positive thing in that light, I suppose: Right now it's not super cheap to run a terrible call center. If five years from now you'll essentially get that for free, companies might see an opportunity to rise above "cheap and terrible" to be more competitive, which would likely create jobs again. Jobs which, however, don't have the main goal of being done cheap and terribly. A way more level playing field than what we have today, where companies compete by chosing the areas in which they want to shine.
Maybe terrible writing, art and customer service becoming nearly free _can_ be a tipping point of sorts. It only wouldn't be if nobody cares, then the market will adjust for that. But I'm not giving up on consumers just yet.
- it allows us to focus on the goals and not the process
- we still need people who understand the PROCESS (developers)
- people who can build tools that bridges the gap between AI generated zero value spam and bespoke human created content.
- we need tools to introduce subtlety and craft to AI outputs
- those are things that are important in a high saturation low margin environment - attention to detail (consider videogame crash of 1980s).
At the moment if i generate image with SD very difficult to change something specific in a nuanced way. that is where tools will emerge made by good developers. high value work is nuanced and often extremely subtle.
An increased ability to introduce subtletly will also open up new niches of interest. because the creator can focus on expressing their ideas in very personal ways.
> There will always be need for top tier leetcoders. but barrier to entry will get much higher.
That's definitely a possibility. The barrier to entry won't be just in terms of how much you need to know, though, but also the fact that society as a whole might not require that many top tier leetcoders.
> There will be an effectively endless demand for anyone who can create 10x the value while only demanding 3x the pay of the typical worker.
That's one possible future.
Another future is one in which a single AI assisted "full tech developer" can solve all the coding problems of a company within say, one week. No company will require the services of an entire "full tech developer", so society will employ roughly one freelancer / consultant per ~30 large enough companies.
I've been interviewing with a lot of companies in the past few weeks. I'll agree that one week is probably optimistic, but most of them seem to have requirements that are in fact pretty simple.
Admittedly, reaching the one week stage might require better programming platforms and APIs.
- looking at AI as "chat" or textbox or AI autocomplete is wrong imo
- companies will come that utilise AI to deliver things faster and cheaper
- you quote $10k and 1 month, we will do it in $1k 1 week
- first to go is low value difficult to automate work which is already offshored and commoditised (basic webdev, graphic design)
- later the middle tier of work, that requires internal context (once whitelabel NDA'ble bespoke AI solutions are mainstream)
----
There will always be need for top tier leetcoders. but barrier to entry will get much higher.