Cape Town, South Africa — 28 Jul 2025
Deep tech and the next decade
We’re halfway through the 2020s, and already the pace of change feels different. Technology is moving out of the lab and into everyday infrastructure. What used to sound speculative now feels surprisingly practical.
Deep tech is no longer a fringe conversation. It’s becoming the foundation beneath how we work, build, and solve.
From AI systems that adapt to biology to Quantum processors that challenge our assumptions about complexity. These aren’t distant possibilities, they’re beginning to influence real-world tools, industries, and decisions. The shift isn’t just in capability, it’s in mindset. We’re finally starting to reimagine the systems around us, not just improve them.
AI, Quantum, and the Evolution of Work
AI in 2025 represents a fundamental shift from reactive to agentic systems autonomous agents capable of planning, reasoning, and executing complex multi-step workflows without human intervention. Unlike traditional AI that responds to prompts, these systems can independently analyse market conditions, develop strategies, and coordinate with other AI agents to achieve business objectives.
Quantum computing has moved decisively beyond the ‘somewhere, someday’ narrative. 2024-2025 marks the transition from theoretical promise to tangible commercial reality. Google’s Willow processor, IBM’s roadmap to fault-tolerant quantum computing by 2029, and Quantinuum’s achievement of ‘three 9s’ qubit fidelity represent concrete milestones in an accelerating timeline.
The way we work is also evolving. Digital twins. Edge intelligence. Virtual collaborators. We’re seeing more responsive systems, not just centralised platforms, but tech that acts at the point of need and the old lines between tech and business are fading. Multidisciplinary, hybrid teams are increasingly the norm, not the exception.
All of this is happening alongside another reality: the climate. Sustainability isn’t a layer to be added later, it’s part of how systems need to be designed from the beginning. From carbon-aware infrastructure to modular, energy-efficient architectures, we’re learning to measure progress differently. Not just by how powerful a system is, but by how responsibly it operates.
The Funding Bottleneck and the Valley of Death
But with all this progress comes a more complicated reality.
The ‘valley of death’ the critical gap between R&D funding and commercial viability remains deep tech’s primary bottleneck. Boston Consulting Group’s 2023 analysis reveals that while deep tech now claims a stable 20% share of venture capital funding (up from 10% a decade ago), the fundamental mismatch between investor timelines and deep tech development cycles persists.
The numbers tell a stark story: deep tech investments take 25-40% longer to mature between funding stages compared to traditional tech ventures, with development timelines stretching 5-7 years before commercialisation. Yet the same BCG research shows that traditional and deep-tech-focused funds deliver similar unweighted internal rates of return (26% and 25%, respectively) suggesting the issue isn’t returns, but patience.
This valley is particularly acute in emerging markets, where weak commercialisation ecosystems and fragmented capital markets compound the challenge. As BCG notes, success requires ‘smarter policies, longer-term capital, and stronger collaboration between public institutions, academia, and industry’ a recognition that traditional VC models need fundamental redesign for deep tech’s unique risk-reward profile.
This isn’t a reason to pull back. It’s a reason to double down, with smarter policies, longer-term capital, and stronger collaboration between public institutions, academia, and industry. Because when deep tech works, it tends to create platform technologies that ripple across sectors, jumpstarting entirely new industries.
“The best way to predict the future is to invent it.” – Alan Kay
Modular Energy and the Deep Tech Stack
Even in energy, one of the hardest sectors to disrupt, we’re seeing foundational change. The fusion race is advancing, but another technology is quietly stepping into the spotlight: modular nuclear power. Compact, efficient, and increasingly deployable, these reactors offer a flexible path to scalable, carbon-free baseload energy. Once the domain of nation-states, nuclear power is becoming modular, localised, and software-defined and powering everything from green hydrogen production to off-grid data centers.
What we’re seeing isn’t a wave of isolated breakthroughs, it’s a deeper shift in how technology, people, and systems fit together. The challenges ahead, from talent to energy to resilience, won’t be solved by any one tool but the growing maturity of deep tech gives us a new kind of flexibility. A chance to design smarter, faster, and more responsive foundations for whatever comes next.
2030 won’t arrive gradually – it’s being built right now through the convergence of technologies that were separate just five years ago. Agentic AI systems managing quantum-optimised supply chains. Carbon-aware digital twins orchestrating energy flows across modular nuclear grids. Edge computing enabling real-time responses that prevent problems rather than react to them.
Toward a New Economic Era
The companies and nations that understand this convergence, that invest in the intersections rather than individual technologies – will define the next economic era. The valley of death that has historically claimed promising innovations is being bridged by new funding models, patient capital, and hybrid public-private partnerships that match deep tech’s unique timeline and risk profile.
This isn’t about adopting the latest technology first. It’s about building adaptive systems that can evolve as the underlying technologies mature. The next decade belongs to those who can orchestrate complexity, not just manage it – who see deep tech not as isolated breakthroughs, but as the interconnected foundation for solving humanity’s most pressing challenges while creating unprecedented economic value.
The next decade won’t be defined by who adopts the latest technology first. It will be shaped by who puts it to meaningful use.