Inteligencia Artificial

General Intuition Seeks $300M at $2B Valuation: Bezos Bets on Spatiotemporal AI

The startup training AI agents in spatiotemporal reasoning is negotiating a round led by Jeff Bezos

June 19, 2026 · 4 min read

an abstract image of a sphere with dots and lines

TL;DR: General Intuition is in talks to raise $300M at a $2B valuation, with Jeff Bezos as lead investor. The startup develops AI agents with spatiotemporal reasoning, key for robotics and autonomous vehicles.

What Happened?

According to TechCrunch, General Intuition is in advanced negotiations to close a funding round of approximately $300 million, which would value the company at around $2 billion. The round would be led by Jeff Bezos, founder of Amazon, through his family office Bezos Expeditions, which has also backed other AI startups like Perplexity AI. The startup, founded by former researchers from Google Brain and DeepMind — two of the world's most prestigious AI labs — focuses on training artificial intelligence agents in spatiotemporal reasoning. This ability allows machines to understand and predict the movement of objects in space and time, a fundamental capability for operating in dynamic physical environments. The information, which has not been officially confirmed by the company or Bezos Expeditions, comes from sources close to the negotiations, so it is subject to change.

Why Is It Important?

Spatiotemporal reasoning is considered the next big challenge in artificial intelligence. While language models like GPT-4 dominate text and code, they lack a deep understanding of the physical world. They cannot predict how an object will move in space, nor plan a trajectory while avoiding obstacles. General Intuition aims to fill that gap, developing systems that can operate in real-world environments, such as robots, drones, or autonomous cars. Bezos's investment signals that the market sees strategic value in this technology, which could be as disruptive as LLMs were for language processing. Historically, AI has gone through several waves: first symbolic reasoning, then statistical learning, then deep learning applied to images and text, and now the frontier is physical interaction. Companies like Boston Dynamics have demonstrated agile robots, but without true spatiotemporal understanding; others like Waymo have made advances in autonomous driving, but with enormous sensorization and annotation costs. General Intuition bets on a learning-based approach, similar to the one that allowed DeepMind to dominate Go, but applied to the real world.

Consequences for the Market and Users

If the round materializes, General Intuition would join a select group of AI startups with multi-billion dollar valuations at early stages. This could accelerate the race for physical AI, where companies like Skild AI (founded by former Carnegie Mellon and Google researchers), Covariant (specialized in picking robots), Physical Intelligence (developing foundation models for robots), and Figure AI (which raised $675M at $2.6B in 2024) compete. Bezos's entry not only brings capital but also a network of contacts and access to Amazon resources, such as AWS and its logistics infrastructure. For users, the impact would be tangible in the form of more autonomous robots in warehouses, home assistants capable of picking up objects, or more reliable navigation systems in urban environments. However, the technology is still in the experimental phase. The path from research to a viable commercial product is long and costly. For example, Boston Dynamics took over 20 years to launch its Spot robot, and it is still not profitable. Moreover, competition is fierce: big tech companies like Google (with its Everyday Robots project, canceled in 2023) and Amazon (with Astro) have tried and failed to commercialize home robots. So, while expectations are high, there are no guarantees of short-term success.

What Readers Should Know

  • The funding is not closed; negotiations could change or fail, as has happened with other high-profile rounds in 2024.
  • General Intuition has not made public statements; the information comes from sources close to TechCrunch, so it should be treated with caution.
  • The $2B valuation reflects high expectations, but also the inherent risk of a technology not proven at scale. In comparison, physical AI startup Skild AI was valued at $1.5B in 2024 after raising $300M.
  • Bezos invests through Bezos Expeditions, which has backed AI startups like Perplexity AI (valued at $3B in 2025) and rocket company Blue Origin.
“Spatiotemporal reasoning is the holy grail of AI for the physical world. If General Intuition achieves significant breakthroughs, it could redefine entire industries like logistics, manufacturing, and transportation.” — Analyst at The Vortiq

Context and Comparisons

General Intuition's round echoes that of other AI startups in 2024-2025, such as Anthropic (which raised $1B at $30B) or xAI ($6B at $40B). However, the focus on physical AI is less common and more complex. While language models benefit from abundant and cheap text data, physical AI requires real-world data, which is costly to collect and label. Companies like Boston Dynamics have taken decades to commercialize robots, while startups like Figure AI have raised large sums but have yet to demonstrate profitability. Bezos's bet suggests he sees a window of opportunity to dominate a niche that could be key in the next decade, similar to his early investment in Google in 1998. However, the macroeconomic context is different: high interest rates have reduced risk appetite, and many AI startups have seen their valuations corrected. For example, Stability AI's valuation fell from $1B to $500M in 2024. Despite this, physical AI continues to attract investment: in 2024, robotics startups raised over $2.5B globally, according to Crunchbase. If General Intuition closes the round, it will position itself as one of the best-funded players in this space, competing directly with academic and corporate labs. Time will tell if its focus on spatiotemporal reasoning is the key to unlocking the next era of AI.

Keep reading