What if we could simulate everything?
What if we could understand dynamic environments and simulate interventions?
The inability to model multi-scale phenomena and their cause-effect mechanisms bottlenecks many promising concepts from ever being prototyped let alone deployed. It turns out we can build the simulation testbeds to efficiently and reliably model technologies and interventions in diverse areas from energy and climate to socioeconomics and agriculture.
What if computation could be reimagined for SCIENTIFIC simulation and AI?
Software development is tightly coupled to the available hardware, and patterns in software have paved the path of hardware architectures for decades. It turns out we can build more efficient, flexible systems with "HW-SW co-design" and "machine programming" brought about by the Simulation Intelligence Operating System.
What if simulators and agents could discover new phenomena?
Existing scientific methods are deductive and objective-based. The next generation of science, brought about by innovations in simulation, AI/ML, and accelerated computing, can be inductive and open-ended; human-machine teams will efficiently explore an exponentially larger space of hypotheses and solutions, in all domains at all scales.
Our aim is to build Simulation Intelligence (SI) technologies and scalable platforms that enable human-machine teams to make Nobel-caliber breakthroughs in key areas of physical and life sciences. We are a new type of venture — a new, hybrid science-focused organization to strike a balance between Bell Labs, SpaceX, Nvidia, and Deepmind (for example) — as our mission is not efficiently achieved through existing models and mechanisms in industry, academia, or government — aka "overedge". To this end we are two "sister" organizations:
Pasteur Labs is a venture-backed, public-benefit corporation for building simulation intelligence technologies from theoretical foundations at TRL-1 to reliable, scalable systems at TRL-9. The name comes from "Pasteur's quadrant" for use-inspired R&D; striking the balance between 'clear problem to be solved' and 'scientific advances required', Pasteur's quadrant defines a class of R&D that calls for ambitious goals that are measurable and testable (in the real world at scale).
Institute for Simulation Intelligence
ISI is a 501c3 non-profit organization to support PhD students, postdocs, and other researchers in open pursuit of simulation intelligence endeavors.
Pasteur & ISI operate out of our headquarters in Brooklyn, NY, while the team is largely distributed (from San Francisco to Boston to London to Berlin).
We are a small team of industry-hardened experts in AI and computational sciences, from leading organziations such as Deepmind, Cerebras, CERN, NASA, and so on. We're structured for highly complimentary skillsets, with specialties that range from multi-physics modeling to cooperative AI to causal inference to machine programming. We strive for interdisciplinary synergies, cultivating a culture that largely emulates Bell Labs, while operating with the nimble, agile best practices of deep tech startups. The unique coordination of Pasteur Labs and ISI helps enable our team to include an expansive group of external researchers at top universities, national labs, and government agencies.
If you share our ambitions and can potentially contribute significant value to the team (in science and engineering, but also in culture, knowledge, and leadership), we want to hear from you: firstname.lastname@example.org. In particular we're hiring for:
Applied ML & Simulation Engineering