AI is a productivity tool. It dramatically increases what a person can accomplish.

This means, for the same output, you need fewer people. If output remains constant, number of people needed will go down.

This is the argument given by lots of people who fear AI will take jobs away in the society.

The thing that’s not being considered here is that AI doesn’t kill competition in marketplace.

If company ABC reduces headcount and their competitor doesn’t, they have higher capacity than ABC, higher scale/capacity than ABC.

With higher capacity, they get cheaper per unit costs, and that in turn helps to either increase margins, or reduce selling price. This, theoretically helps you take higher marketshare than your competition.

The other company will either crumble due to competition, or they will hire as many people as they can afford to reduce their costs.

Also, why would a company want to produce the same output when they can leverage AI to get higher outputs. I mostly have experience with software companies, and they’re always tied up in bandwidth. They have more work than they’re able to complete!

The nature of the jobs will change, as some of the current jobs will not even need a single person to complete that work, but there will be a need of someone to use the AI tool to get the job done, and thus we come back to number of humans being the rate determining step in the whole thing.

Of course, the whole argument breaks down if you assume that we’re close to upper ceiling of market activity, but we’ve seen in the past that increase in productivity also increases the addressable market.

AI and humans are both leverages to use. If there’s demand, it makes sense to use both the leverages rather than just one.

So, will AI lead to higher unemployment? No.

Will it lead to jobs changing in their function? Yes.