Bezos' Prometheus Startup Raises $12B to Create 'Artificial Engineer'

Jeff Bezos unveiled details of a new AI startup Prometheus, which has already attracted $12 billion in investment. The company is developing a software product called "Artificial General Engineer" (AGE) — a tool for designing, testing and manufacturing physical products, as well as for developing pharmaceutical drugs.

What is Artificial General Engineer

Prometheus is creating software of a fundamentally new class: not a narrow-purpose tool, but a fully-fledged "digital engineer" capable of taking a product from initial idea to finished product. The key distinction from generative language models — AGE works not with text, but with matter: drawings, loads, specifications, technological maps and manufacturing constraints. Bezos's conceptual bet: the physical world remains underequipped with AI. While language models help with writing and programming, designing an aircraft engine, synthesizing a new material or developing a drug candidate molecule still requires years of work by teams of engineers. AGE is meant to change this.

The announced functionality includes:

• generative design — AI creates constructive solutions from given constraints on weight, strength, cost
• virtual testing: load tests, thermodynamics, aerodynamics without a single physical prototype
• manufacturing process optimization for specific equipment and factory
• molecular design in pharmaceuticals — search and selection of promising chemical compounds
• creation of digital twins for predictive testing across all lifecycle stages

Target clients of Prometheus are industrial companies and pharmaceutical giants for whom reducing the engineering cycle from 10–15 years to 2–3 means not just convenience, but competitive advantage and tens of billions in savings.

Scale of the round and competitive landscape

$12 billion at launch — one of the largest initial rounds in the history of technology startups. For comparison: Anthropic reached comparable total funding figures only after several years of active work, and OpenAI built its investment portfolio gradually starting in 2019. Such a volume of capital from the start speaks to a long-term horizon: physical AI does not deliver quick returns, but the industrial R&D market is measured in trillions of dollars.

Industrial AI is being pursued by major technology players. Google DeepMind researches the application of neural networks in materials science and biochemistry, NVIDIA builds digital twin infrastructure through its Omniverse platform, Microsoft develops AI copilots for engineers in collaboration with Siemens and other industrial partners. However, no full-fledged "AI engineer" covering the entire engineering cycle is yet in production use by any of them — and it is precisely this gap that Prometheus intends to fill.

Pharmaceuticals as the second front

A special place in the company's plans is occupied by drug development. The traditional path from molecule to pharmacy shelf takes 10–15 years and costs $1–2.5 billion. The AlphaFold breakthrough has already shown the potential: predicting the three-dimensional structure of a protein, which previously took years, now takes hours. AGE is positioned as the next step: from structure prediction — to the complete cycle of drug development accounting for efficacy, toxicity and synthesis manufacturability. The most complex part is regulatory: the FDA, EMA and other agencies are still only formulating requirements for AI-generated drug candidates, and Prometheus will enter this segment at a pivotal moment.

What it means

Prometheus is a clear signal of a shift in AI investment: money is going not into the next chatbot, but into "physical intelligence". If AGE delivers even part of what is promised, it will reformat the economy of industrial R&D and accelerate the launch of new drugs to market. $12 billion at launch is not just a large round, but a public declaration of intent to lead in industrial AI for the next decade.