Leadership in the Age of AI

In the midst of all the hype-driven talk about which jobs AI is “inevitably” going to replace, I have recently been thinking about the concept in more specific terms. I say it a lot and I stand behind it fully, the idea that AI capabilities shouldn’t be replacing any human roles, but instead be replacing various key processes within a role. That’s what will open up human capacity in a way that grows businesses and amplifies individual career trajectories.

The middle tier of any business is the easiest level to analyse in this regard. AI frees up middle level workers to accomplish higher level tasks. The mundane and repetitive parts of their work can be streamlined and delegated to AI, allowing the human workers to take on more judgement based roles, be it oversight or growth or management.

What can potentially happen at the entry levels of a business is admittedly more problematic and worth a deeper dive as well, but for different reasons than what I’m getting at with this tangent. In short, it’s easy to see when AI takes the routine work off the junior roles, you also take away the ground people used to learn on. The busy work was never just busy work for those roles, it was the apprenticeship, essentially. That’s a serious issue and it’s a leadership issue, not something to hand to HR and forget about. But as I said, that’s a tangent for another day.

Those reflections got me thinking about what happens at the top, then. If the entry level roles are hitting the ground running (for better or worse), and the middle tier is freeing up their capacity in a way that leverages them up to judgment based roles, what then happens to the people at the top? Do they get lifted right out of the rotation and onto mega yachts to live out the rest of their prime? Surely not, and yes I’m being intentionally hyperbolic.

What actually happens at the top, from what I’ve observed, is much more interesting. For most of business history, the key challenge for leadership was getting a clear picture of your industry and your mission. Good information was slow and expensive, and a lot of what senior roles were paid for was that they could see the situation more clearly than the competition, for whatever reasons. The leaders with the best reads “won.” But now, in so many ways, AI takes that advantage and hands it out to everyone at the same time. Your people can now research a decision to a depth that used to take a full team a month, but now by end of day.

This is what causes the bottleneck to move up. It used to sit at “can we figure this out.” Now it sits at “we’ve figured out five good ways to do this, so which one is the best one.” It’s a much harder problem to work out, but an exponentially more productive one, if you have the right people at the helm. When you’ve got one decent option to consider, the decision makes itself. But when you’ve got five, each one well researched and genuinely sound, more analysis won’t break the tie. The analysis is what made the tie.

Somebody still has to choose. And the choosing is the part no machine and no amount of research can do for you. You pick one path out of five good ones, you commit, and you carry the why. And it has to be a “why” that you actually believe in, because the moment something unexpected hits, a choice made for a real reason can bend and a choice made by coin toss just breaks. The buck stops harder now than it used to, precisely because everything leading up to the buck got easier.

Making a choice like that is also carrying the people. The team below you is staking its work on your call and the investors are staking their money on it, and they need to be able to follow the reasoning, or it isn’t leadership, it’s just a guess made by the person with the biggest office. So much focus has been on worries about AI taking too much work off the bottom of the pyramid, but what’s really interesting is considering how much more exposed it makes those at the top.

It’s all pretty reminiscent of when CRM landed on sales floors. The pitch was that the system would track every account and every stage of the pipeline, and for a while people assumed that was the end of the salesperson who stayed on top of it all. But it wasn’t, of course. The reps whose whole value was just remembering who to call were in some trouble. But the ones who could look at the pipeline the system handed them and say which of those deals were actually real, and stake their time on it, became worth more than ever. The tool took over the record keeping and pushed the human up into the judgment zone. AI is doing that now, only this time it’s running the move on the whole organisation at once, all the way to the top.

Every technological leap moves the bottleneck somewhere new. AI isn’t eliminating the need for people. It’s simply moving the hardest part of every role further up the value chain. And at the top of the organisation, that means leadership becomes less about having better information, and more about having the courage and conviction to choose what to do with it.