Where to draw the line between what AI does and what people keep
The most important decision CEOs will make in the next few years isn’t how fast to adopt AI. It’s where to draw the line between what AI does and what people keep.
To better understand this decision, I spent part of my summer vacation reading “Machines of Loving Grace” (Dario Amodei), “Human Life in a Post-AGI World” (Tyler Cowen) and “A World Without Work” (Daniel Susskind).
Susskind argues that cost pressure will push companies to replace as many people as they can, and governments will tax the winners to fund a universal basic income for the rest.
I understand this logic but I think companies that follow this path risk being quickly commoditized.
Amodei says meaning comes mostly from relationships and connection rather than economic labour. For most adults, those relationships get built at work. The colleague you mentored. The customer who trusted you when it mattered. The team you got through a hard year. The team dinner when you won a big deal.
Take away work and you take away what most of us find meaningful. Me included. As much as I love football, I don’t want to spend all day watching it (especially as I support Chelsea).
Meaningful work is why the best people join you and why they stay. Companies that automate purely on cost will get the same AI as everyone else and lose the one thing competitors couldn’t copy.
So rather than rushing to replace human labour, don’t we need to redesign every role? For every job, draw the line: this is what AI does, this is what the human keeps. I’d draw that line even where AI can do the task at least as well, because the point isn’t only output. It’s the judgement, trust and commitment people build by doing it.
The obvious things to keep are the human parts. Building trust. Making people feel heard. Showing up.
But I’d go wider. Taste: AI can generate a hundred options, but deciding which one is right for this customer, or simply us, is a human call. Judgement: the decisions where the data runs out. Accountability: a name on the decision. Purpose: AI can tell you how to get somewhere, not where you should want to go.
For a CEO like me, that means the people side: hiring, motivating and mentoring. And owning the decisions that shape what kind of company we are. Not the analysis behind them, where AI will out-think us and we should let it.
For a teacher, it means coaching and mentoring students, and the judgement of when to push and when to back off. Not the grading, the corrections, or explaining the same concept to 30 students at one pace. Hand that over and give teachers the time to actually know their students.
So the human side of our jobs, often the most meaningful part, is about to become the whole job. Companies that get this right will build something durable. Those that treat it as a pure cost exercise will look back on it as a strategic mistake, not just an ethical one.
If you run a company, what would you keep for the people who work there?
Originally posted on LinkedIn.