@CSB It's not that black and white, and just cause almost every company is on the LLM bandwagon doesn't mean they can actually make a good use of it. Just like a self taught software developer could automate their job at a library to not have to work anymore exists somewhere, it is more likely 'that guy' just went to work as a software developer somewhere in the first place. There maybe was a time in 2000's when this was more likely, but its almost non-existant now.
Heres a bonus contender to your options:
C) People who can master 'AI' will work trying to implement these tools in companies. Companies will see some efficiency gains from it, but in general the bulk of implementations will either kinda suck and not do what the business actually needs, fall short of expectations and require additional work to be useful or be borked by average employee in their daily use. Just like ms word/excel became a 'standard skill' at office job, so will some 'ai tard wrangler app' will become a new standard. Some current work or inefficiencies will phase out, to be replaced by new set of inefficiencies and work.
So yeah, the market will shift to a new skill set over time, but I don't think its gonna be that much of a paradigm shift like horses and cars, it's just that all of the current hype is very wide/visible.
It's just nobody really noticed that some chicken meatpacking plant has had vision based machine learning systems that accurately weigh, sort and identify any deficiencies in poultry for last 10+ years now.
And while we do have the opportunity now for a marketing person to automate themselves if they pick up this new cool skill, it will be gobbled up eventually, as most economically viable opportunities do, and life will go on. We will just look back at 'x' role like we look at switchboard operators today.
I'm a bit broken record on this, but see 'nanotechnology hype' to see the likely path AI will take in the future (I agree that AI is a bit more impactful, but the general idea stands)
