Tech • AI • Robotics • Game
The 5 revolutions that will create the fortunes of tomorrow
7/10TL;DR
Artificial intelligence, robotics, energy, deeptech health, defense and space are emerging as the main engines of wealth creation, with the biggest risk seen as remaining a passive user rather than an investor.
KEY POINTS
AI has moved beyond hype
ChatGPT reached 1 billion users in four years, faster than Facebook, which took six. The boom is no longer framed as pure speculation because demand is already producing revenue, margin expansion and infrastructure shortages across chips, electricity and computing power. In the United States, roughly one-third of recent business-related financial growth is attributed to AI or AI investment.
The AI value chain extends far beyond chatbots
The market is structured in layers: applications such as legal assistant Harvey; model providers such as OpenAI and Anthropic; cloud and neocloud operators supplying compute; then chipmakers and energy suppliers underneath. This stack has created multiple profit pools, from software interfaces to server racks, cooling systems and data centers, making the winners broader than a handful of consumer brands.
Infrastructure suppliers are capturing outsized gains
Hardware and infrastructure have become central beneficiaries. Dell, long known for personal computers, now derives about 40% of its business from AI-related products such as servers and cooling systems. Nvidia has multiplied in value as demand for AI chips surged, while compute customers are reportedly paying about 40% more for capacity than they did six months earlier because supply remains tight.
Tokens are being treated like refined energy
A growing view in the sector is that AI output, especially tokens, should be understood as a refined form of energy. The comparison likens raw electricity or fuel to the underlying resource, and tokens to the usable product delivered to end users. The argument is that, just as economies rarely complain about having too much energy, demand for tokens may keep expanding as businesses automate more intellectual work.
This cycle differs from the dot-com era
Unlike the early 2000s, when fiber networks were built ahead of real usage, current AI expansion is described as demand-led. Revenues are already material, margins can be high, and bottlenecks are visible across semiconductors, power supply and data-center capacity. Anthropic was cited as generating about 80% gross margin on inference, highlighting that some model businesses are already throwing off significant cash before training costs.
Robotics is emerging as the next wave
The next step after digital intelligence is physical execution. Physical AI combines models with robots that can manipulate objects, navigate factories and perform repetitive labor. One projection cited by Elon Musk envisions 20 billion robots globally in the coming years, implying a market potentially larger in unit terms than cars and one that would reshape logistics, manufacturing and household services.
Energy is becoming a strategic choke point
AI and robotics both intensify electricity demand, pushing investors toward nuclear, grid capacity and power contracts. Some AI-focused cloud operators are signing electricity purchase agreements for 30 years to lock in future supply. The logic is straightforward: if compute is the engine of AI, energy is the binding constraint, and any technology that unlocks more power could become economically decisive.
Deeptech health is attracting serious capital
In medical technology, Chipiron illustrates why deeptech is being watched closely. The company is developing a new generation of MRI systems designed to be lighter, cheaper and more accessible. After producing mostly laboratory images two years ago, its machines are now generating in vivo images on humans, a milestone that shows how difficult hardware can be but also how large the upside may be when breakthroughs arrive.
The US retains a structural lead over Europe
Three advantages stand out: capital depth, a large single domestic market and a more aggressive entrepreneurial culture. US founders can raise tens of millions rapidly, scale nationally across 300 million consumers, and operate in an environment that often tests technology before regulators catch up. Europe, by contrast, is seen as constrained by fragmented markets, smaller pools of private capital and more cautious regulation.
The main investment lesson is to own the picks and shovels
The strongest opportunities may lie less in famous consumer apps than in enablers: chips, neoclouds, data centers, energy systems, vertical AI software, imaging suppliers and robotics data providers. Public markets still offer exposure, but much of the value creation now happens before listing. The result is a widening gap between private-market upside and what retail investors can buy once companies are already worth billions.
CONCLUSION
The central bet is that the biggest fortunes of the next two decades will come from backing the infrastructure and platforms behind AI, robots, energy and deeptech, not merely using their products. In that framework, the costliest mistake is not choosing the wrong trend, but standing aside while the value is created.
Ask a question