AI’s impact on the “mothers of all natural sciences”, theoretical physics and mathematics, will likely be profound and fundamental. For AI represents a new form of intelligence, not merely a higher intelligence. The commercial potentials, however, will be difficult to capture.
This blog post is the nineth in the series on AI’s disruption of innovation
This blog entry is the nineth in a series about the disruptive impact of a technology based on a new form of intelligence that is self-learning, universally enabling, and allows for deep customization:
- Where does AI impact today?
- What is new and disruptive about AI?
- AI’s effect on robotics and experimental automation
- AI’s effect on climate and meteorological research
- AIs impact on social sciences
- AIs impact on education and mental health
- AI’s impact on biotechnology and pharmaceuticals
- AI’s effect on materials research and quantum chemistry
- AI’s effect on theoretical physics and mathematics
- What should investment strategies in AI take into account?
A New Form of Intelligence Will Lead to New Fundamental Insights
AI will affect our epistemological understanding of the world, because it represents a new form of intelligence, not merely a faster intelligence. That fundamental realization will influence all other research.AI will therefore particularly affect the methods of the natural sciences, i.e. how we think, explore, and prove. In itself, the commercial potential is thus limited, but overall this may be where the greatest breakthroughs will arise.
The potential is vast. In reality, we only have certainty about the 5–10% of the universe that our senses can register. The rest is dark, meaning it does not interact with light. The only thing we truly know about the universe, therefore, is that it can be described and projected quite precisely with mathematics.
Epistemological breakthroughs from AI will thus particularly affect research methods , i.e. hypothesis generation, simulations, discoveries, and proof..
At First, It Will Especially Affect the “Mother Sciences” of Physics and Mathematics
Concretely, this has implications for astro- and nuclear physics. The volume of datasets from, for example, particle accelerators and cosmological observations is today overwhelming. This makes it difficult to see real or conditional correlations. It also makes it difficult to derive logical connections, i.e. to achieve genuine understanding.
- AI’s raw computational power combined with its pattern recognition has already had success in finding structures in quantum fields and condensed matter. This points toward new theoretical frameworks.
- Structurally, AI has proven effective at creating surrogate models (approximations) for complex systems, which drastically reduce computation times.
- Gemini 2.5 (DeepThink), for instance, has produced breakthroughs in mathematics because it applies multiple approaches to problems and then cross-checks the internal logic. Humans often work in the same way.
- In 2021, Sydney University used AI models based on DeepMind to find new patterns in representation theory. These have since been proven by humans.
- AlphaGeometry and AlphaProof have proven effective in generating proofs within formal mathematics and geometry.
- In addition, AI assistants have already shown themselves to be effective as “mathematical partners,” adapting to researchers’ individual working styles and suggesting methods of proof, etc.
- Google DeepMind (the basis for Alpha), Caltech, and Anthropic are the largest AI providers in the field. Furthermore, the EXCLAIM project in nuclear physics is among the dominant systems.
Overall for the Series of Blog Posts, It Will in the Short Term Be the Big AI Names That Dominate. Seed, However, Holds Great Potential
Taken together from the previous blog posts in this series, it is reasonable to expect the following disruptions on investment horizons:
Short term (0–3 years): Hyperscalers and chipmakers will be in the strongest positions, though competition among existing players is intense.
- There are high entry barriers in most areas. This rewards companies such as NVIDIA, TSMC, AMD, Intel, and ASML, as well as cloud platforms like Microsoft, Google, Meta, and Amazon. All of the latter are currently developing their own AI chips. Huawei, Xiaomi, and Cambricon may also play larger roles over time.
- Materials, pharma, biotechnology, mental health, education, and climate technologies will receive increased seed focus. In general, there are major disruptive opportunities at TRL 1–4 levels.
- Several development breakthroughs will be so significant and happen in such a short span of time that they will be unsettling. This will increase overall volatility among AI companies.
- There will be such major advances toward AGI that many developers will claim to have the first. Whoever has the first has the best chance of becoming the largest, and thus the most valuable. That in itself could carry geopolitical significance.
But in the medium and long term, large companies may perish, and new giants may emerge
Medium term (2–6 years):Here, new and disruptive software layers will appear, and industry-specific AI agents will become widespread. Deep customization will become a basic condition for many companies, and this will affect their cost structures. The value-creating stages in many products will shift. “Slow adopters” may face difficulties, even if they are large and influential with their customers.
- Materials, pharma, biotechnology, mental health, education, and climate technologies will see breakthroughs and many unicorns (IPOs above USD 1 billion).
- AGI will mean that the pace of technological development will begin to outstrip what many specialists can keep up with. It will also create periods of uncertainty and misunderstanding. Wrong strategies may lead to “Intel” and “VHS” outcomes for many former crown jewels.
Long term (+5 years): Materials research, biotech, and quantum-augmented AI will have grown so large that they can shift power balances. Fundamental research, i.e. mathematics, physics, quantum chemistry, etc., may deliver breakthroughs on the scale of the discovery of electricity.