Sources: Eric Schmidt is building an organization, modeled on OpenAI, to tackle scientific challenges by using AI; most of the funding comes from his wealth
Reed Albergotti / Semafor : X: @simonkhalaf and @reedalbergotti Forums: Hacker News and Slashdot X: Simon Khalaf / @simonkhalaf : The best person to do this. Congrats @ericschmidt https://www.semafor.com/... Reed Albergotti / @reedalbergotti : New scoop in today's @semafor tech newsletter: Eric Schmidt has an ambitious new project to create the “OpenAI of science” - still in the early stages and not public yet. But could be a big deal. (also this caption pun by @GinaChon is amazing) https://www.semafor.com/... Forums: Hacker News : Ex-Google CEO Eric Schmidt to Launch AI-Science Moonshot Msmash / Slashdot : Ex-Google CEO Eric Schmidt To Launch AI-Science Moonshot
Context & Ripple Effects
This report extends Schmidt’s earlier AI investment push, including investments tied to US-China AI competition, from backing companies to building an organization around a specific application: scientific work.
The scientific-AI focus also foreshadows the later launch of FutureHouse’s AI-scientist tools and API, making this an early signal that well-funded AI efforts were moving beyond general-purpose models toward research workflows.
First-order effects
- Schmidt can assemble and fund an AI-for-science organization without relying primarily on outside investors, giving the still-private effort room to set its own research agenda.
- Researchers and technical hires focused on applying AI to science gain another potential employer and funding source modeled on the frontier-lab approach.
Second-order effects
- Existing AI labs and science-focused startups face additional competition for specialized researchers, model talent, and partnerships with scientific institutions.
- A prominent, personally financed entrant raises the incentive for other AI backers to frame frontier-model development around concrete scientific applications rather than general capability alone.
Third-order effects
- If such efforts produce useful research tools, AI-for-science may develop into a distinct lab category: capital-intensive model organizations organized around domain expertise and scientific outputs.
- The pattern would further concentrate agenda-setting power in a small group of wealthy patrons and major labs, even where the stated goal is broad scientific progress.
The trend: This is one early example of frontier-AI funding shifting toward purpose-built organizations that promise scientific and strategic applications.