Periodical workshops He came out of Stealth on Tuesday with a $ 300 million war breast as a round seed, supported by a technological industry that Andreessen Horowitz, DST, Nvidia, Accel, Elad Gil, Jeff Dean, Eric Schmidt and Jeff Bezos.
The magazines were founded by Ekin Dogus Cubuk and Liam Fedus. Cubuk drove the team of materials and chemistry to Google Brain and Deepmind, where one of his works was, for example, an AI tool called Gnome. This tool discovered over 2 million new crystals in 2023Materials that could be used one day to supply new generations of technology, the researchers say.
The Fedus is a former Vice President of Openai and one of the researchers who helped create Chatgpt. Also drove the group that created the Neurosurgery Network of First Trillion Parameter.
His small team is also full of researchers who have worked on other AI science and materials from the building of the Openai agent to work at Microsoft’s Mattergen, an LLM Materials Science Discovery AI.
The goal of the journal laboratories is nothing less than the automation of scientific discovery, creating AI scientists, the company says. This means that robots carry out natural experiments, collect data, repeat and try again, learning and improving as they go.
The first goal of the laboratory is to invent new superconductors who hopes to better deliver and possibly require less energy than existing superconducting materials. But the well -funded start hopes to find other new materials.
Another goal is to collect all the data of the natural world produced by AI scientists as they mix and heat and otherwise handle various forces and raw materials in searching something new.
TechCrunch event
Francisco
|
27-29 October 2025
“So far, AI scientific advances come from internet -trained models” and LLMS “exhaust” the internet as a source that can be consumed, the company says in an introductory blog. “[A]T magazine, we build scientists AI and autonomous workshops to work. ”
The hope is that not only the workshops will invent next generation materials, but will produce invaluable fresh data that AI models can consume to continue their evolution.
While this may be one of the most impressive teams of researchers to assemble a start for this purpose, it is not the only one working for AI scientists. AI as a tool for automating the discoveries of chemistry was Subject of academic research from at least 2023. It is the pursuit of tiny newly established businesses such as Tetsuwan Scientific, as well as non -profit organizations Future house and the University of Toronto Acceleration consortium.
