Something important has changed in the venture capital landscape, and most people in the industry have not yet fully reckoned with it. After a decade in which the dominant investment thesis was that any industry could be disrupted by a team of talented software engineers and a modest amount of capital, that playbook is running out of runway. The easy wins have been won. The digital layer of commerce, communication, and entertainment has been built. What comes next requires something far more difficult: genuine scientific and engineering innovation at the level of matter, energy, and biology.
This is the era of deep technology. And it represents, in my view, the most important opportunity in venture capital since the internet.
The End of the Easy Arbitrage
To understand why deep tech is now the compelling frontier, it helps to understand what happened during the software era. The period from roughly 1995 to 2020 was characterised by a simple and powerful dynamic: the world had built an enormous amount of analogue infrastructure - shops, banks, newspapers, taxis, hotels - that could be dramatically improved by moving it onto software platforms. The venture opportunity was, in a sense, an arbitrage. You did not need to invent new technology. You needed to apply existing technology - web servers, mobile applications, cloud computing - to legacy industries that had not yet adopted it.
This was genuinely transformative. It created enormous wealth and made many aspects of everyday life dramatically better and cheaper. But the arbitrage opportunities are increasingly exhausted. Every major industry that could be served by a software layer now has multiple competing incumbents. The marginal return from building the next food delivery app or the next HR management platform is low. The valuations required to win in crowded software markets are high, and the competitive moats are shallow.
The next wave of venture-scale returns requires something different. It requires companies that are doing things that were previously impossible - not merely more convenient or cheaper. Companies that are creating genuinely new capabilities rather than redistributing existing ones.
Why Deep Tech Creates Better Businesses
There is a common misperception among generalist investors that deep tech is inherently riskier than software and therefore demands a lower allocation of capital. This perception is understandable - hard technology companies typically take longer to reach commercial scale, require more patient capital, and face a category of scientific risk that software companies do not. But the perception misses something crucial about the nature of the competitive landscape once a deep tech company succeeds.
Software businesses are famously easy to copy. The same talent pool that built your product can build a competing product. Your competitive advantage depends on network effects, switching costs, and brand - none of which are guaranteed. When a large incumbent decides to replicate your product with ten times the resources, your options are limited.
Deep tech businesses are different in kind. A company that has spent five years developing a novel catalyst formulation for hydrogen production, or a new memory architecture for edge AI inference, or a proprietary mRNA delivery system for targeted therapeutics, has built something that is genuinely difficult to replicate. The scientific knowledge is embedded in the people, the processes, and the patent estate. It takes years to accumulate. Competitors cannot simply hire a team and rebuild it in six months.
This means that the moats in deep tech tend to be durable in a way that software moats often are not. When a deep tech company wins its market, it tends to win it for a very long time. The return profile, for investors who can exercise the patience required, is extraordinary.
The Three Megatrends Driving Demand
One of the most important differences between deep tech investing and software investing is the nature of the demand. The demand for deep tech solutions is not driven by consumer taste or the adoption curve of a new application. It is driven by fundamental structural forces that are irreversible and growing.
The first is the energy transition. The world has made a binding commitment to decarbonise its energy systems over the next thirty years. This requires not only new generation capacity from wind and solar, but entirely new infrastructure for storage, grid management, long-distance transmission, industrial heat, and transportation. The total capital investment required is measured in the tens of trillions of dollars. Every component of that transition represents a potential market for deep tech companies.
The second is the compute constraint. Every major advance in artificial intelligence is running into the same problem: there is not enough computing power, and the computing power that exists is too expensive and too energy-intensive. Solving this requires innovation at every layer of the computing stack - processor architectures, memory technologies, interconnects, compilers, and cooling systems. The companies that succeed in removing or expanding the fundamental physical constraints on computation will be among the most valuable in the world.
The third is the transformation of healthcare. The convergence of genomics, proteomics, advanced materials, and machine learning is creating the conditions for a fundamentally new model of medicine - one that is predictive, preventive, and precisely targeted to individual biology. The incumbent pharmaceutical industry, for all its resources, is poorly structured to capture this transition. The next generation of therapeutic and diagnostic companies will be built from scratch, by founders who understand both the biology and the technology.
The Talent Shift
A structural change that I believe is underappreciated is the shift in where the best talent is choosing to apply itself. For the past twenty years, the most ambitious and technically capable graduates of the world's leading research universities largely flowed toward careers in software engineering, often at large technology companies or consumer internet startups. Physics PhDs took jobs at hedge funds. Chemists became data scientists. Biologists pivoted to machine learning.
This is changing. The generation of scientists and engineers entering the workforce now has grown up with a different set of reference points. They have seen the climate crisis unfold. They have watched a pandemic expose the fragility of healthcare systems. They have lived through a decade in which the most powerful AI systems in the world hit fundamental physical limits. And they have seen, in the extraordinary story of mRNA vaccines, an example of what can happen when basic science reaches the scale of industrial deployment.
More of the best technical talent is now choosing to work on genuinely hard problems. This is creating the founding teams that deep tech venture investors need to back. Five years ago, finding four co-founders who had all done deep work in, say, solid-state battery science and also had the ambition and risk tolerance to build a venture-backed company was genuinely difficult. Today, it is common.
What This Means for Investors
For investors, the deep tech opportunity requires a different set of capabilities than traditional software investing. The most important is scientific credibility. A venture investor who cannot engage substantively with a battery researcher's claims about electrolyte chemistry, or evaluate a quantum computing company's coherence time data, is at a fundamental disadvantage. They cannot perform meaningful due diligence, they cannot add value to portfolio companies, and they cannot distinguish between founders who understand the limits of their science and those who do not.
This is why we built Lumino Capital the way we did. Our investment team includes practising scientists and engineers who spent years in research before turning to venture. We have the ability to engage with technical claims at the level of the underlying physics and chemistry. We can call out the promising from the implausible, the rigorous from the wishful. This is not a nice-to-have in deep tech investing. It is a prerequisite.
The other requirement is patience. Deep tech companies do not move on the timelines of software startups. A battery company that is building a working prototype at the seed stage may take three years to reach commercial pilot, and another two years to scale manufacturing. This is not a failure of execution - it is the nature of the work. Investors who push for artificial acceleration of these timelines tend to destroy more value than they create. The right answer is to set accurate expectations from the beginning, fund the company with enough capital to reach genuine technical milestones, and resist the pressure to optimise for short-term signalling.
The European Advantage
There is one final point that I believe deserves more attention than it typically receives: the exceptional position of European deep tech. Europe has a world-class basic science infrastructure, built over decades of public investment in research universities, national laboratories, and pan-European research programmes. The Max Planck Society, the Fraunhofer Institutes, CERN, the European Space Agency, IMEC, the Wellcome Sanger Institute - these organisations are generating foundational scientific advances at a rate and quality that rivals any in the world.
What has historically been missing is the commercialisation infrastructure to translate that science into companies. The venture ecosystem in Europe has been smaller, more risk-averse, and less willing to back genuinely early-stage technical companies than its American counterpart. This is changing rapidly, and the gap between the quality of European science and the maturity of European venture creates a genuine opportunity for investors who are willing to do the work required to identify and back the best science coming out of European research institutions.
At Lumino Capital, we have spent six years building relationships with research groups across Europe's leading scientific institutions. We have developed a sourcing capability that we believe is genuinely differentiated. And we are confident that we are at the beginning of a very long runway of compelling investment opportunities in European deep technology.
The deep tech decade is not a prediction. For those of us who spend our days in the company of the scientists and engineers working on the problems described in this article, it is already apparent. The question for the venture industry is not whether this wave is coming, but whether traditional investors have the tools, the talent, and the temperament to ride it. We built Lumino Capital in the conviction that the answer, for us, is yes.