The usual speculation I hear here is that top-level jobs are likely to be the first to go. It’s usually in a context that highlights AI’s abilities. Never in a context that highlights cost cuts. However, the idea that top position employees could form a 'good-old-boys/girls’ club to exempt themselves from being replaced for budget reasons seems absurd. Furthermore, today I cringed when I saw a blurb that listed Data Scientist as in danger, but Dishwasher as safe. It doesn’t get any clearer than that. Who cares if washing dishes is labor intensive. A dishwasher is cheap. 10,000 financial analysts aren’t.
In my opinion and according to my observation, rather than a job removal, it will be a reduction in the workforce. A dishwasher can be fully replaced by a robot, which does repetitive dishwashing, but jobs that require more mental effort would seem to still have a supervisor.
The term that has stuck with me is not “reduction,” but “augmentation.” Such as one employee can be augmented by AI or machinery. Yours may be more realistic though. Ultimately, I think corporations will present themselves in the most advantageous way they can. I presume that “reduction” would be too blunt, and “augment” too evasive.
Something that stands out is that AI is not alone. History is filled with advancements that have reduced workforces. Still, AI’s potential is hard to put into perspective. Can anything in history be compared to the Staregate project?
That’s the thing, though - these technologies haven’t reduced workforces. The world, including the US, still has a high labor participation rate (though, it’s slowing in some countries due to (what I believe to be) demographic challenges). Unemployment is low, but productivity has continuously gone up. In my humble opinion, I think that’s what we’re more likely to see in the coming decade or so, is much more being produced with more efficiency, rather than a decline in jobs.
I do have to agree. If your job is anything that is a clerical, administrative or data entry kind of job, then AI is a threat to your role. It will probably reduce the workforce, with those that remain moving to more of a supervisor role. This is because LLMs can do these tasks under supervision but do need human correction at times. This also includes a lot of entry-level jobs for recent graduates, as the roles they performed are replaced mostly by LLMs. If you look at the job market nowadays, good luck.
However, if your role requires you to be customer-facing or if replacing you requires a large amount of robotics, then you are probably secure for a while. This involves job roles such as warehouse work or retail, especially if you are multiskilled. Likewise, professional roles such as Doctors are secured, and it will mostly be AI augmentation as these roles cannot be done independently with our current models. Also, if your role is in the armed forces, that will also increase in size, mostly due to an increase in military funding.
If you are in a creative industry, it’s more complicated. You may get replaced by AI for cost-related reasons. However, since these industries thrive on human creativity currently, it may keep your role in place. Independent artists will face more competition due to AI, but they will still exist. Likewise, if you are self-employed, you are safe in your role.
However, LLMs do create this dangerous entry-level gap between low-skilled work that is hard to automate at this time and high-skilled work. It is going to cause an employment crisis at the high skill level, as graduates will not have the skills to enter the high skill level immediately and thus will be underemployed for their role. Without graduate roles, graduates will not be able to gain a foothold in their industry. You have to pick the mitigating factor carefully to deal with this. Unemployment may be low at this time, but underemployment is rampant. The question is thus, how do we replace the entry-level/mid-level roles people need as a career?
I think that businesses are eventually going to have to face the music - they will run out of developers as seasoned developers retire, and they’ll have to hire the new-grads. I think of it more like an increased training cost, but then again, even the new grads don’t have to focus as much on the little stuff; they can get right into learning what the senior developers (and equivalent, for other fields).