Innovation is no longer about technology, it's about execution

Q: In the age of AI, is innovation still a competitive advantage or has it simply become a prerequisite? 

In the software industry, innovation meant building technology complex enough to create barriers to entry, what we call a “moat”. That required years of R&D, highly specialised engineering teams and, in some cases, patents. 

That model is now fading. AI is accelerating experimentation at scale, lowering the barriers that once protected innovation. And that’s a good thing: smaller players can now challenge established leaders. 

AI is pushing us into a velocity-driven market. When everyone has access to the same tools, innovation alone stops being a differentiator. What matters is how fast you turn it into something real. That’s exactly what we’re doing at VirtualBrain: turning the power of AI into plug-and-play solutions for professional services. 

Q: Is AI still a driver of differentiation, or is it standardizing innovation? 

I don’t think AI is standardizing innovation itself. For now, humans are still in the driver’s seat. What it’s standardizing and accelerating is the innovation process by drastically reducing the cost of experimentation. 

Today, a team can explore a market or build a prototype in a few hours. Failure costs almost nothing. 

That’s the real shift. Companies can learn and iterate far more quickly than ever before and redirect their energy where it really counts: understanding customers, testing ideas, building better offers. 

Q: How would you define “useful innovation” in today’s business environment? 

There’s a mantra you often hear in entrepreneurship: real innovation is the kind that genuinely improves the lives of the people it’s designed for. It sounds obvious, but when I launched my first company, I learned just how easy it is to fall in love with the technology itself. 

We had built an incredibly sophisticated financial forecasting engine. On paper, it was brilliant. In reality, nobody wanted to buy it because we hadn’t started with a real business problem. That’s a common trap in deep tech: building showcase technology that’s technically impressive but disconnected from real-world needs. 

Useful innovation is, above all, innovation that creates value. It comes in two forms: incremental innovation, which improves existing products or services, and disruptive innovation, which fundamentally changes the way things are done. 

Q: Many companies are multiplying AI POCs but struggling to scale. What makes the difference? 

AI pilots succeed or fail because of people. The same use case can deliver completely different outcomes depending on who’s leading it. You need credible business champions, people who can build momentum and bring teams along. 

Then there’s a second mistake I still see too often: starting with the tool. The starting point should always be the problem, not the platform. 

And finally, scaling only works if the groundwork is solid. Start small. Test with a focused group. Refine the use cases. Prove the value. Then expand. 

Q: What are the most common mistakes you see in AI projects today? 

Probably the biggest one is the search for the “perfect AI tool”—as if there were a single solution capable of doing everything. AI is becoming a new operating system built on an ecosystem of specialised tools: some designed for product development, others for software engineering, design and UX, customer service, and more. 

Another common misconception is the belief that a two-line brief can lead to a study worthy of publication by McKinsey. That completely overlooks the reality of corporate work: methodologies, validations, workflows, and human and organizational constraints. 

Q: If you had to pick one KPI to measure the success of an AI initiative, what would it be? 

Time saved. In practice, you need to look at three dimensions: time saved, output produced, and quality delivered. 

A successful AI project is one where a team can say: “What used to take us five days now takes one, and we’re delivering more, at a higher level of quality.” 

The real question is what you do with that time. AI isn’t about reducing headcount. It’s about freeing up capacity and redirecting it toward higher-value work: innovation, customer relationships, new services and strategic advisory work. 

Q: Will innovation remain fundamentally human, or will it increasingly be delegated to AI agents? 

AGI (Artificial General Intelligence) promises near-infinite intelligence at almost zero cost. In that world, humans could theoretically become the bottleneck. 

But I don’t see a fully autonomous economy taking over. Humans will stay in the loop, not just for oversight, but because trust still depends on human relationships. 

What I see emerging instead is the role of the “agent manager”. Tomorrow’s workers will orchestrate fleets of AI agents, delegating much of the execution while staying responsible for judgement and direction. 

Today, we use AI. Tomorrow, AI will rely on us to decide. 

Q: Looking ahead, what major shift do you believe will have the greatest impact on businesses over the next few years? 

I often come back to a quote by Warren Buffett: “Only when the tide goes out do you discover who’s been swimming naked.” 

I think AI will play exactly that role in the information economy. For years, many businesses have created value simply by acting as intermediaries or by offering a better user experience. AI is now beginning to absorb some of those layers. 

The companies that thrive won’t be those with the most spectacular AI. They’ll be the ones that turn it into practical, integrated solutions that solve real business problems. 

Don’t forget your swimsuit !