In a week, on 1 and 2 November, we open the third edition of AI Expo Europe in Bucharest, at the Radisson Blu. I am writing this during the week when everything comes together: lists, emails, final checks, phones ringing at odd hours. It is the moment when, as an organiser, you have less time to think than ever. Which is exactly why I wanted to stop for an hour and write.
There is another reason. This is article number 100 in the series I have been writing here. It felt right for the last one to be about the subject that has kept me busiest in recent years, and about a question I hear more and more often: fine, the excitement is behind us, so what now?
Three editions, three different conversations
When I wrote about why we founded AI Expo Europe in January 2024, the idea was simple: Central and Eastern Europe needed its own meeting place for business and AI, and Bucharest could be that place. I was not alone; we started it with a team that has worked incredibly hard ever since.
The first edition was about building something from scratch and about basic questions: what can AI do, what can it not do, is it just a passing wave? I wrote honestly about that in our first-year review. By the second edition the conversation had already moved on: almost nobody asked “what is it” any more, but “how do we put it into production” and “how do we measure whether it works”. In what I learned at AI Expo Europe 2025 I described how the bottleneck was no longer the technology, but the people.
The third edition arrives at a moment when AI is no longer new to anyone. It is in your phone, your inbox, your accounting software, your search results. And when a technology becomes ordinary, that is when the serious questions start.
After the hype come the hard questions
Hype has its uses. It brings attention, money and talented people. But it has a cost too: it makes many buy promises instead of results. The phase in which every company announced “we have integrated AI” just so as not to fall behind is, I think, largely over. The questions I hear now from founders and managers are much more down to earth and, frankly, much healthier:
- What value does it actually bring to my company? In money, in time, in quality.
- How do I implement it without breaking what already works?
- Who is accountable when it gets something wrong? What internal rules do I need?
- What happens to my people? Whom do I train, and for what?
- How do I keep customers’ trust when part of the interaction is handled by a machine?
Value, implementation, governance, people, trust. If I had to sum up what comes after the hype, those would be the five words.
What we want to put at the centre this year
I will not share the agenda here, or any names. I will only tell you what we set out to do as a team, as an intention, when we planned the 2026 edition across its two stages, the conference stage and the demo stage:
- Real implementations and results. Fewer visions of 2035, more stories about what was done, what worked, what did not, and why.
- Agents in production. AI agents have been talked about a lot as the next step after chatbots. We care about what happens when they actually start working inside a company’s processes.
- AI, jobs and skills. Not in an apocalyptic tone, but practically: what changes in people’s roles and how you prepare them.
- Regulation in practice. The AI Act (the EU’s law on artificial intelligence) is no longer a theoretical topic. Companies want to know what they need to do, not just what the text says.
- Trust and safety. Because without trust, any implementation stops at the first incident.
- The region’s role. Central and Eastern Europe should not just be a consumer of technology built elsewhere.
These are intentions. How well we turn them into good conversations is for you, the people who come, to judge.
From GPT-3 to today: my journey on this blog
Rereading the archive, I realised how much of what I have written here has been about AI. Back in 2020 I was marvelling at a GPT-3 text that sounded human. Two years later I wrote about my first weeks with ChatGPT, with the enthusiasm of someone who has just found a new toy and suspects it is not just a toy.
Then came the articles on how I use it in business, on GPT-4 and the fear that jobs would disappear, on European regulation, on agents, on AI companions and their ethics, on who is accountable when the algorithm gets it wrong. Read one after another, they look like a diary: from amazement, to experimentation, to questions about responsibility.
At the start of this year I also wrote 10 predictions for business in 2026. I will not grade them yet; the year is not over. But the broad direction, AI moving from experiment to everyday operations, is visible everywhere around me.
What I have learned as a founder in three years
First: an event is not about the stage, it is about what happens between sessions. The best projects I have seen get started began in the hallway, over coffee, not at the microphone.
Second: the audience matures faster than you think. What was exciting two years ago now feels introductory. If you do not grow with it, you lose it.
Third: good questions matter more than quick answers. With AI, anyone can give you an answer in seconds. What is missing is someone who helps you ask the right question for your business.
Let’s talk in person
If you are in Bucharest on 1 and 2 November and you care about the practical side of AI, come along. Not just for the talks, but for the conversations. Find me in the hallway. I love hearing what people are trying in their companies, what worked and especially what did not, because that is where you learn the most.
And if you cannot make it, write to me. These conversations do not end when the event does.
The blog stays
This blog has always been the place where I write down what I learn: from business, from mistakes, from travel, from technology. The series ends here, but the writing does not. I will keep putting here what I find out, including what comes out of the third edition.
Until then, a question for you: if you had to pick a single AI topic to discuss seriously, without the hype, what would it be?