When speed becomes a commodity

There was a time when being fast was enough. In the startup world, speed became almost a value in itself, to the point where we measured a company's competence by how quickly it shipped. Whoever got there first pulled ahead, built the platform earlier, gained six months of advantage and used that time to dig in while the others were still building.

In the second episode of Life After AI we talked with Diogo Dutra, founder of CAOS Focado and a venture builder who has worked with frontier technology for years, about what happens when that math stops adding up. And it is stopping, because AI turned those months of advantage into a week. If anyone can build fast, and now anyone can build fast, then getting there first no longer protects you.

Over the years, venture capital turned the path to success into a formula. First you become software, then a platform, then you need network effects, and so on. These formulas gained the status of thesis and of cleverness, and an entire generation of founders learned that success only existed if you followed that script. The problem is that a formula is precisely the kind of thing anyone can repeat. When the technical barrier falls, the advantage that came from the formula falls with it.

When building gets easy, it becomes tempting to skip the hardest and least valued part of all, which is discovery: finding out whether the problem you are solving is actually worth solving. Diogo said that in his work with deep tech he does not have that choice, because getting the direction wrong costs three years of research. That forces a kind of care that the rest of the ecosystem, used to pivoting cheaply, has been unlearning. He prefers to call it intensity rather than speed, and I think the distinction matters, because you can keep a delivery rhythm without turning haste into an end in itself.

The conversation passed through Incorruptible, Eric Ries's new book, and through an idea that ties the rest together. We tend to treat quality and ethics as things that compete with results, as if caring for them meant giving up efficiency. But the reading Ries seems to propose, and that we built up along the way, is that in the long run these vectors converge. Focusing on quality and keeping your values in place is not the slow path; it is the path that builds the company that lasts, with customers who stay and a brand that means something. The shortcut is what tends to be a flash in the pan.

If we think of an organization as an organism, the biggest challenge in adopting AI today may not be the technology but the lack of formalized knowledge. Companies do not stop to reflect on their own memory. And memory here is not a stored archive; it is something alive, consolidated yet in need of constant reassessment. AIs project the past very well and invent futures that are linguistically coherent, but not necessarily truly coherent. Formalizing an organization's memory does not make people dispensable. On the contrary, it puts each of them in a continuous process of reflecting on what the company knows and on what has already aged.

If you work in innovation, invest, or are trying to make AI work inside your organization, I would very much like to hear how you see this. Has speed become a commodity where you are too? And what do you think became valuable in its place?

The full episode, with Diogo, is on the Life After AI YouTube and Spotify. The episode is in Portuguese.