AI and real-world visual understanding remain unsolved. Machines can recognize a face, caption a photograph, describe a scene, and outperform radiologists on narrow diagnostic tasks. So, if machines can see what we see, they must be doing what we do. They don’t. Nature does not produce straight lines, perfect circles, or right angles. AI lacks comprehensive human visual reasoning.
Artificial intelligence is no longer a tool that the life sciences industry is adopting. It is a force that is relocating where value is created and who captures it. Three costs are collapsing at once: drug discovery, company independence, and the ability to reach the patient. Each of these costs was, for forty years, a moat protecting the incumbents who could afford to pay it. In addition, the capacity to discover and manufacture medicine has become a strategic infrastructure in the same category as energy, semiconductors, and compute. The molecule has become computable. The architecture of value creation in life sciences has changed.
Albert Camus’s warning from nearly 80 years ago, that humanity is subordinate to abstraction, people are replaced by calculations, and the willingness to accept suffering as an administrative variable persists. We have industrialized the human crisis. We are at an inflection point where the consequences of our choices, both good and bad, will arrive faster, hit harder, and spread more widely than any prior moment in history. We have the proven capacity to recover from previous crisis. The question is whether the next crises potentially makes recovery impossible.
AI models produce raw intelligence. They generate tokens. But tokens are an intermediate good, not a finished product. What customers actually pay for is legal work completed, code shipped, claims processed, research synthesized, and decisions supported.
They pay for refined output.
Attention has focused on the infrastructure layer — the frontier labs, the compute stack, and the data centers. That attention is not misplaced, but it overlooks a structural shift already underway. Once you understand the model as an intermediate good rather than the end product, the center of gravity moves. The decisive question is no longer who can produce intelligence, but who can turn it into something usable, trusted, repeatable, and economically defensible.
In other words, who can refine it into a usable product?
At the base of the chain sit the token producers — OpenAI, Anthropic, Google DeepMind, Meta, DeepSeek, and Qwen. They produce raw capability. This layer is expensive to build, technically formidable, and still moving fast. But crude oil is not gasoline.
Enterprises and consumers pay for gasoline.
Software Is the central nervous system of the global economy and its demise is greatly exaggerated. There’s a growing narrative thatsoftware is becoming commoditized. Large language models write code. Autonomous agents assemble applications. The barriers to building digital products appear to be collapsing. If software can be generated instantly, then software itself must be losing value.
This conclusion fundamentally misunderstands how technological disruptions develop and expand. Software is becoming the infrastructure layer of modern civilization. The economic, industrial, and geopolitical systems being constructed over the next three decades will not run on software. They will run as software.
Space is no longer a frontier. For most of the modern era, space has been misunderstood—not technologically, but economically.
Space was a destination rather than a system and a heroic engineering challenge rather than an industrial platform with a continuous operational, and commercial potential. Many early “commercial space” narratives sought to impose venture logic on a domain that remained structurally dependent on government capital, prestige economics, and one-off missions. The result was predictable: excitement without durability, valuation without cash flow, and ambition without a stable market. Now, space is about economic persistence: building businesses that treat space not as a product but as a technological and economic stack – a physical layer supporting a stack of software services and networks.
As the era of artificial intelligence is here, it’s easy to fall into the trap of despair and fear over the loss of control and the worry that artificial intelligence is about to unleash killer robots and enslave humanity. Either that, or Artificial intelligence will improve lives, expand access to education, advance healthcare, and advance climate science, among many other improvements. Luckily, AI’s benefits greatly outweigh its costs. Nothing is free, and everything comes with a price (there are always both sides to the ledger), but the extraordinary benefits that artificial intelligence can unleash are worth the effort. It would be a mistake to slow down, pause, or restrict research, development, and AI applications. AI will not destroy the world — and it is more likely to save it.
The wheel was a great invention. But not until it was combined with other wheels to create a usable cart was it an innovation. The wheel was a breakthrough; a moving, stable cart was a system. Systems create intelligent, scalable, and disruptive technology. Innovations are not new technologies. Breakthroughs are necessary, but it’s systems that are the solution. The value created by AI in the physical world is not scaling software. It is focus, discipline, and constraint within effective systems. The systems that endure will not be those that promise universality, but those that dominate specific economic niches, involve humans strategically, and survive year ten of operation.
Robotics and related technology are ready for deployment, but the industry hasn’t crossed the threshold into full-scale production. Computational breakthroughs in stunning demonstrations are attention-grabbing, but the realities of industry quickly take over. There is a gap between robotics and artificial intelligence (“physical intelligence”) as it transitions from potential to hardware delivery in a demanding industrial setting. Physical AI and its integration into robotics may become one of the largest markets in history. But it is an industrial problem whose solution is not on a software timeline. In other words, its commercial deployment requires much more systems integration and real-world constraints than a software slide deck contemplates.
Markets destroy the comfortable assumption that tomorrow behaves like yesterday. They reward those who can identify when the system’s structure changes and punish those who try to fit new realities into old frameworks.
That is why the conventional idea of “what something is worth” has become less relevant than how systems evolve. Investors who cling to formulas intended for stable conditions will always be surprised by nonlinear disruption.
Nowhere is this more obvious than in AI and energy, where the variables are not just changing; the equations themselves are being rewritten.
