Life After AI: the beginning of a conversation

Life After AI is a series of conversations about what changes in our lives and our organizations in a world transformed by artificial intelligence. It grows out of a conversation that had been going on among us for months, in which we discussed what these tools amplify, what they erode, and what we still do not understand about them. At some point we realized, and found the courage to admit, that this discussion should not stay confined to us. Not because we had reached every conclusion, but because the questions had matured enough to be worth asking in public.

The idea behind Life After AI is to publish recorded conversations on a regular basis, with no fixed script and no pretense of delivering ready-made answers. In each episode we start from a study, a case or an unease, and examine the subject calmly, including the disagreements among us, which exist and are part of what makes it interesting. We also want to bring in guests who live close to the topic and whose reflections can enrich the conversation.

That is also the reason for this text: we want to talk with people who are thinking about or working on these questions. If you lead a team that has adopted these tools, research the subject, or have simply asked yourself what you should or should not delegate to an AI, write to us. Disagree, bring a case, point us to someone who should be part of the conversation.

Full episodes come out every two weeks on YouTube and Spotify. We publish clips on YouTube and Instagram, and texts like this one on LinkedIn and Substack.

Where we begin: the augmentation trap

The first episode starts from a recent MIT study, The Augmentation Trap, by Michael Caosun and Sinan Aral. Its main finding is striking: the adoption of AI in organizations can be a trap. At first, augmentation delivers what it promises, and people produce more, faster and with less effort. But continued use, when not accompanied in the right way, displaces precisely the practice through which competence is built and maintained. Over time the bill arrives, and what is left are people dependent on the tools, with less capability of their own than they had before adopting them.

What the conversation showed us is that it is not a single trap but several, nested inside one another.

The first is the most counterintuitive: falling in can be a rational decision. The study's model shows that even those who understand the mechanism and anticipate the erosion tend to adopt anyway, because the gains arrive now and the costs arrive later. Knowing about the trap protects no one from it.

The second hits those who are just starting out. When the tool replaces judgment instead of complementing it, experienced professionals keep developing, while the more junior ones delegate exactly the practice that would have formed them. The study shows trajectories that diverge permanently: some realize their full potential, others become dependent on the tool. If today's juniors do not walk the path that forms a senior, who will be the seniors in our organizations ten years from now?

The third is about incentives. Managers and directors evaluated on quarterly results are drawn to the quick gains that AI promises, and the decision is rational for them, because the benefit shows up during their tenure and the cost shows up after they have left. Those who carry that cost are the professional whose competence has worn down and the organization that has lost capability.

And there are quieter traps. The measurement trap: productivity has stopped being a good signal of competence, and almost no organization measures whether people are getting better or worse at what they do. And the trap of degrading use: the vigilance of the early days, when we still review and question what the tool delivers, gives way to passive acceptance, and an adoption that began healthy slides into the trap without anything in the design having changed.

The episode develops each of these paths and also the other side: the traps are not inevitable, and there are ways of adopting AI that preserve and even exercise people's competence. We will not give everything away here. Watch the full episode on the Life After AI YouTube or Spotify. The episode is in Portuguese.

The technology is not going back into the bottle. The question that interests us is not whether we will live with it, but how.