AI does replace people (and thank goodness)
Resources will be abundant, ideas will be scarce.
When I was running the family advisory firm, a cloud-based invoicing software company came to see us. They told me they were there to help us be more efficient. I told them that yes, they would help improve efficiency, but at the expense of the advisor. They denied it, but once they saw I was fine with it, they admitted that, sooner or later, many clients would no longer need an advisor for their bookkeeping and tax filing. They were right.
Many advisors tried to deny the obvious, pointing at the mistakes the software was making back then. But the invoicing software was soon joined by an invoice reader that booked expenses automatically, and a reconciliation tool that handled bank transactions on its own. The change kept accelerating, and today there are tools that do all of this at the click of a button, with a precision close to 100% in many companies.
That is why, when I hear people defend that Artificial Intelligence will not replace humans, I think back to what happened in my office almost 15 years ago.
If we have to say that AI replaces people, we say it, and that is that.
The arrival of AI in several industries is already triggering mass layoffs and a sharp drop in junior hires, and this will only accelerate from here. As with automated bookkeeping, the technology is not perfect, but it gets better every day, and in many cases it already outperforms what people do.
Does that mean unemployment will skyrocket?
Look at it from a different angle: how often have you said in recent years that there is a shortage of qualified people? And the business owners you talk to?
With shrinking demographics across the world, AI not only can replace people, it must do so, if we want our companies to have the talent they need, human or otherwise. We are heading toward an era where resources will be abundant, and where what will truly be valued is ideas: it is not a question of whether you will have enough people to do what you want, but of thinking carefully about what you actually want to achieve. And that is far from easy.
A business owner who works with two AI agents that he supervises shared, a few weeks ago in a meeting with other owners, that he had never felt more overloaded. The two agents produce more work than he can possibly absorb. The bottleneck has moved up: it is no longer a lack of capacity to execute, but a lack of capacity to decide well what is worth doing.
Let me ask you the same question I have asked my clients, and that none of them has answered yet: if you could hire a person who would do exactly the work you wanted, done right, what would you have them do?
© Oriol López Villena 2026