And without expertise you end up asking stupid things to a token producer machine, however godlike it can be.
It's hard not to extrapolate...
I have noticed myself a 3 months or so ago sometimes asking AI questions for things that I knew the answer to (but had to think about) and my solution would have been better. Once I realized it I realized how despite always telling my self to think about the result I was still allowing myself to hand over some of my thinking to the AI.
What I have started doing is still doing lots of things without AI and trying to not lose those skills because I think they are still needed but AI can lull you into handing over your prefrontal cortex, which is scary.
Humanity can then relax in floating chairs drinking soft drinks and consuming AI generated entertainment.
AI won't end silicon valley, but could still shrink it significantly.
In which case your sentiment is probably correct. And now that I think of it, this is potentially very bad news for the outsourcing companies in countries like India. They also have less money to spend on tokens.
"give me all the specs in a list" -> "which architectural changes would simplify the implementation based of these specs" -> delete everything -> "implement the new specs"
I don't think there's going to be a rush to outsource. In fact, I suspect that it's quite the opposite scenario going to play out.
Primarily because now that the cost to produce the boring/mundane stuff that is the stuff of outsourcing (internal tooling, dashboards, ERP, integrations, QA etc.), it will likely be much more cost effective to hire 3 devs in the US and give them hefty token budgets and have much easier coordination and communication loops than outsourcing to a Bangalore body shop that will try and bill you 15 engineers for the same work.
In fact, word on the street here is that there's a significant correction in the Bangalore realty market for the first time in history thanks to the IT salaries drying up. [1]
I suspect LLMs are going to have a global worldwide deflationary effect what with white collar salaries at risk. The only counter acting force might be the distortions caused by recent wars and tariff shenanigans of the US government.
But even having said all of that I am hopeful for the eventual benefits human beings can reap from the advent of LLMS. The only thing that keeps me up at night is my lack of faith in the aging gerontocracy (worldwode phenomenon) that all our futures are in the hands of that seems to be ill equipped to handle this period of what is most likely massive upheaval and uncertainty.
[1] https://timesofindia.indiatimes.com/real-estate/news/housing...
Exactly, tight E2E test coverage and obviously manual functional verification to seal the coverage gaps will be needed even more.
Functionality iterations or adaptations are just faster with good architecture and principles.
In building/construction industry, certainly one might expect delivery speed to matter, and humanity have literally millennia of experience in the domain, and yet it’s notoriously known to go through the roof all the time.