Three Big Ideas #28
Some options on stocks, income illusions, and an aria to innovation
Welcome to our weekly Three Big Ideas roundup, in which we serve up a curated selection of ideas (and our takes on them) in entrepreneurship, innovation, science and technology, handpicked by the team.
🇪🇺 Philip Salter, Founder
My colleague Eamonn was in Brussels yesterday for the launch of When Europe Scales, a very timely report offering practical policy recommendations to grow the European startup and scaleup ecosystem.
Today, I’ll touch on one of the 16 headline policy recommendations – to take a harmonised and competitive approach to stock option taxation:
“Retaining talent also requires a competitive tax environment, particularly in terms of the uniform treatment of stock options across member states. Harmonised taxation criteria across the EU are needed, and this should focus on taxing stock options at the time the stock is sold, not when the option is exercised. EU-Inc could provide the right framework to deliver this coordinated effort at EU level.”
This is a huge deal for startups. As last year’s EU Startup Nations Standards report set out, recognising stock options as capital rather than income, as well as avoiding double taxation are significant challenges. It’s also important to see harmonisation around the ability of startups to offer no voting rights, without which too many people can become involved in decision-making.
These ideas are downstream of work undertaken years ago when founders banded together to make the case for better stock option policies. But this is all taking too long. If anything good is going to come of Trump’s tariffs and broader turn against America’s long-term allies, it needs Europe to break down the regulatory barriers preventing a truly single market. Tear down these walls!
💷 Eamonn Ives, Research Director
For anyone invested in Britain’s economic future, last month’s growth forecasts made for sobering reading. Worse still, some were quick to note that our already modest projections for economic output will be spread across a growing population, due to continued immigration. That sounds like a double blow – low growth, diluted further by more people.
But this framing deserves closer scrutiny. Yes, higher GDP per capita is generally better than lower GDP per capita – but it’s a crude proxy for changes in individual welfare. It’s influenced not just by how productive we are, but by who we count. When immigration increases the denominator, it’s easy to miss the effects on the overall composition, productivity, and dynamism of the economy.
Consider this: when someone moves to the UK to work as a cleaner or delivery driver, they don’t necessarily compete with the British-born lawyer or doctor – rather, they often complement them. By enabling higher-productivity people to work longer or more efficiently, lower-wage workers can boost total output. That’s comparative advantage at work, even if on the face of it GDP per capita goes down.
With this in mind, there’s a case for looking beyond GDP per capita to something like GDP per ‘native’ capita. I’ll be the first to admit, the phrasing makes me feel a little squeamish. But if immigration increases total output without dragging down natives’ earnings, that seems worth noting. After all, if what’s really driving the lower average is compositional change, not declining prosperity for existing residents, then some of the gloom may be misplaced.
It won’t resolve every debate, but it might just help us distinguish between real economic problems and statistical artifice.
🚀 Anastasia Bektimirova, Head of Science and Technology
How can we design aircraft that can reliably find thermals and wind shear, harvest energy from them, and interlace these into long, unpowered flight paths? As we explore the ocean as a new frontier, what high-value foods and materials could we sustainably cultivate? This is the flavour of questions that will be explored by the newly announced second cohort of Programme Directors at the Advanced Research and Invention Agency (ARIA). What makes this cohort different is that ARIA deliberately sought out entrepreneurial scientists who have successfully built ventures, communities and technologies that have left a mark on society. I enjoyed reading about the selection process here.
Each Programme Director is joining ARIA with one or two early areas of exploration, that they will shape before narrowing down and defining programmes more precisely. There is a lot to be excited about – from engineering biological energy to sculpting innate immunity. But I’ll be watching two of the emerging programmes particularly closely.
First, the Collective Intelligence Engine that could revolutionise how scientists navigate the research landscape. Discoveries often emerge from identifying new connections between studies across different disciplines. What if AI could map every scientific argument, instantly revealing valuable connections, contradictions, and gaps? This is what the programme aims to do by building a living knowledge engine. But beyond this, it leaves room for ARIA-shaped metascience questions. For example, how might AI restructure scientific disciplines, and what institutional changes would be needed once this happens? The answers to such questions could be as transformative as the technology itself, reshaping how we organise, fund, and evaluate scientific work in the age of AI-assisted discovery.
Second, the Extending Our Perception programme that focuses on computational systems that can simultaneously process multiple streams of data from advanced sensors far beyond what humans can perceive. This could transform healthcare by detecting diseases before symptoms appear, enable more precise environmental monitoring, or advance sustainable food production. Regular readers will remember that I wrote about this direction earlier, noting how multi-sensor integration would create strategic advantages by gaining a fundamentally different understanding of the world around us. This direction is one of those that could serve as a backbone for many more new ones – just imagine AI systems capable of processing multiple sensory and more conventional data streams simultaneously, revealing patterns and relationships that remain invisible to traditional analysis methods.