To future co-owners
How a swarm of AI agents challenge mathematics
A swarm of 10 000 AI agents have now generated a reported solution to the Navier-Stokes Millennium Prize Problem, a problem which has remained unsolved for nearly 90 years. This is a shocking illustration of how AI can challenge even the purest of sciences, mathematics.
The reactions from the mathematicians have been diverse, ranging from excitement and curiosity to skepticism and concern.
ThermoPhys develops Lean for process engineering
In mathematics, AI agents can work with Lean, a formal programming language that allows mathematical proofs to be verified by computers. But what about engineering?
One of the main reasons why swarms of AI agents have not yet revolutionized how very expensive engineering teams develop more efficient, cheaper and safer processes is that they don’t have the right tools.
This is exactly what we are developing at ThermoPhys. We are building the layer that gives the AI agents access to the laws of thermodynamics (including the second one!), balance equations for mass, energy, charge and momentum, and the most accurate thermophysical models developed by mankind, all connected through a graph of mathematical models that accurately represent real process technology.
Why ThermoPhys?
AI agents are only as capable as the tools they can access. If thermodynamic calculations are inaccurate, fail to converge, and are too slow, AI agents cannot reliably use them to develop industrial processes.
Solving advanced thermophysical property models accurately, robustly, and computationally efficiently requires highly specialized expertise. There are very few people in the world with world-leading expertise in this field. At ThermoPhys, we have brought several of them together.
This expertise has allowed us to give AI agents access to thermodynamic models that reproduce experimental data within measurement uncertainty, while solving the underlying equations robustly and efficiently. Combining these capabilities with process simulation and AI creates an opportunity that is exceptionally difficult for others to replicate.
TP-Process is already being used by paying customers, and one AI agent helps them to build and solve industrial process simulations in TP-Process. We are now preparing for the swarm.
A swarm of AI agents building process technology
Soon, a swarm of AI agents will start building the next-generation process technology with ThermoPhys. In fact, we are incredibly excited to see what they will come up with!
Perhaps they will:
- Reinvent cement production, making CO2 capture unecessary?
- Figure out how to turn sunlight, seawater and air into clean fuel, fresh water and fertilizer, cheap enough for us to turn entire deserts into forests?
- Discover an entirely new industrial process that no engineer has ever considered, because nobody had the time?
Some of these ideas might be fiction, others could change entire industries. We are building the tools to find out. The question for you is ↓