17th June 2026 - Tamas

I want to make a habit of writing reflections.

It is something I found incredibly beneficial in one of my previous roles at Boba: taking the time to write down what I was learning, what I was wrong about, what seemed obvious in hindsight, and what still felt uncertain. This post is part of my commitment to do that again. Hopefully, it will also be useful to others working in a similar space and wrestling with similar questions.

The guiding principle behind Isolio has been shaped by a quote from Jeff Bezos that I have kept coming back to since starting this venture in 2024:

"I very frequently get the question: 'What's going to change in the next 10 years?' And that is a very interesting question; it's a very common one. I almost never get the question: 'What's not going to change in the next 10 years?' And I submit to you that that second question is actually the more important of the two -- because you can build a business strategy around the things that are stable in time" — Jeff Bezos

That question, “what is not going to change?”, feels especially important in AI.

Right now, most conversations are understandably focused on what AI will be able to do next. Will agents replace workflows? Will software interfaces disappear? Will SaaS survive? Will humans still be the primary users of business software?

On that last question, I think the answer is already becoming clear: SaaS will not be used by humans in the way it is today.

In many ways, that change has already started. The old model of business software was built around humans logging into tools, navigating interfaces, filling in forms, moving records between systems, interpreting dashboards, and manually coordinating work across a growing number of applications. That model is already being reshaped by AI. Interfaces are becoming thinner. Workflows are becoming more automated. More of the work that used to happen through clicks, searches, tabs, and dashboards is beginning to happen through agents.

A lot of software will become invisible. Many of the interfaces we use today will collapse into the background. The job of software will be less about giving humans a place to manage work, and more about enabling work to be done.

But the disappearance of today’s SaaS interface does not mean the disappearance of the underlying business problems.

Businesses will still need to win customers, serve them well, understand risk, make good decisions, coordinate teams, manage exceptions, and build trust. They will still operate in messy environments where context matters, trade-offs are real, and the “right” answer is not always obvious.

So the most important question is not whether AI will become more capable than humans at many individual tasks. In many areas, it already is.

The more interesting question is whether AI on its own will be more capable than AI paired with a human.

Our bet at Isolio is that, in the highest-value business problems, the answer will continue to be no.

That does not mean humans stay in the loop because AI is weak. It means humans stay in the loop because the work that matters most often requires more than execution. It requires judgment, taste, accountability, empathy, commercial intuition, and an understanding of context that is often scattered across people, relationships, incentives, and history.

Agentic AI will dramatically change how work gets done. But the businesses that win will not simply be the ones that automate the most. They will be the ones that understand where to combine autonomous systems with human judgment in a way that produces better outcomes than either could produce alone.

That is the principle we are building around at Isolio.

We are not trying to preserve SaaS as it exists today. We believe that version of software is already changing beyond recognition. Instead, we are trying to understand what business software becomes when agents can do more of the work, humans can focus more of their time on judgment, and the interface between the two becomes the product.

SaaS may disappear from the user’s screen, but the business problems it was built to solve will not.

This is still an early hypothesis. I expect parts of it to change. But writing it down matters because it gives us something to test against reality.

And that is what I want these reflections to be: a record of what we believe, what we are learning, and how our thinking evolves as the world changes around us.