# Karol Hausman

> Source: https://aiwiki.ai/wiki/karol_hausman
> Updated: 2026-06-28
> Categories: AI Companies, People, Robotics
> License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)
> From AI Wiki (https://aiwiki.ai), the free encyclopedia of artificial intelligence. Reuse freely with attribution to "AI Wiki (aiwiki.ai)".

**Karol Hausman** is a roboticist and entrepreneur who is the co-founder and chief executive officer of [Physical Intelligence](/wiki/physical_intelligence), a San Francisco startup founded in 2024 that builds foundation models for [robotics](/wiki/robotics). Before founding the company he was a staff research scientist at [Google DeepMind](/wiki/google_deepmind), where he helped create the SayCan, RT-1, and [RT-2](/wiki/rt_2) robot-learning systems, and he was an adjunct professor at [Stanford University](/wiki/stanford_university) who co-taught its deep reinforcement learning course. [1][2][3][4] He holds a Ph.D. in robotics from the University of Southern California. [5][13]

Physical Intelligence, often styled "Pi" or with the Greek letter "π," develops [vision-language-action models](/wiki/vision_language_action_model) intended to control many kinds of robots across many tasks, and it released the π0 (pi-zero) and π0.5 models in 2024 and 2025. [1][2][6] As of its November 2025 funding round the company was valued at about 5.6 billion dollars and had raised more than 1 billion dollars in total. [16][17]

## Who is Karol Hausman?

Karol Hausman is a robot-learning researcher turned founder. He spent roughly a decade working on how robots can acquire general-purpose skills, first as a Ph.D. student and then at Google, before leaving in 2024 to co-found and lead [Physical Intelligence](/wiki/physical_intelligence). [1][3][16] His central thesis is that the scaling behavior seen in [large language models](/wiki/large_language_model) might also apply to robot control if a single model is trained on enough diverse data across many robot bodies and environments. [1][3]

## What is Karol Hausman's education background?

Hausman studied mechatronics at the Warsaw University of Technology in Poland, where he earned a B.Sc. and an M.Sc. [3][13] He went on to a second master's degree in robotics at the Technical University of Munich in Germany, completing a thesis on object segmentation and recognition using interactive perception. [3][13]

He then moved to the United States for doctoral study at the University of Southern California. There he worked in the Robotic Embedded Systems Laboratory under the supervision of Gaurav Sukhatme, and he completed his Ph.D. in 2018. [5][13] His doctoral research dealt with robotic perception and control, including interactive and active perception, where a robot uses its own physical interaction with objects to perceive them better. [3][5] During this period he also collaborated with researchers connected to Stefan Schaal's robot learning group at USC. [5]

## What did Karol Hausman work on at Google?

After his doctorate, Hausman joined Google, working first in the Google Brain team and later in [Google DeepMind](/wiki/google_brain) after Google merged its research groups. As a staff research scientist he focused on [robot learning](/wiki/robot_learning), [reinforcement learning](/wiki/reinforcement_learning), and the use of large models for robot decision making. [3][4] His stated research interest was enabling robots to acquire general-purpose skills in the real world, with a growing emphasis on building and using foundation models for robotic control. [3]

Hausman was among the authors of SayCan, a 2022 project also titled "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances." SayCan paired a large language model with a set of learned low-level robot skills so that a mobile manipulator could carry out long, abstract instructions given in natural language. The language model proposed useful steps, and learned value functions estimated which of those steps the robot could actually perform in its current situation. [4][7] According to Hausman's professional profile, the SayCan paper received a special innovation award at the 2022 Conference on Robot Learning. [3]

He also contributed to the Robotics Transformer line of work. RT-1, published in 2022, was a transformer model trained on a large set of real robot demonstrations to perform many manipulation tasks from camera images and text commands. [4][8] [RT-2](/wiki/rt_2), released in 2023, extended this idea by co-training on web image and text data together with robot data, producing an early vision-language-action model that could transfer some web-scale knowledge into robot behavior. [4][9] This research connected to broader efforts at Google such as PaLM-E and the [Open X-Embodiment](/wiki/open_x_embodiment) dataset, a shared collection of robot data gathered across many laboratories. [9][8]

Alongside this applied work, Hausman published earlier research on multi-task and self-supervised robotic reinforcement learning, [imitation learning](/wiki/imitation_learning), and skill discovery, studying how a single robot could learn many behaviors and reuse them. [3] His Google Scholar record lists thousands of citations across these areas. [3]

## Did Karol Hausman teach at Stanford?

While at Google, Hausman held an appointment as an adjunct professor at [Stanford University](/wiki/stanford_university). [3][4] He co-taught CS 224R, the university's course on deep reinforcement learning, which covers methods for training agents to act through trial and error and reward. [3] His teaching and research overlapped with the work of Stanford faculty in robot learning, and two collaborators from that environment, [Sergey Levine](/wiki/sergey_levine) and [Chelsea Finn](/wiki/chelsea_finn), later became his co-founders at Physical Intelligence. [10][11]

## When did Karol Hausman found Physical Intelligence?

Hausman co-founded [Physical Intelligence](/wiki/physical_intelligence) in 2024 and became its chief executive officer. [10][11] The company emerged from stealth in March 2024, and its founding group included Hausman, [Sergey Levine](/wiki/sergey_levine), [Chelsea Finn](/wiki/chelsea_finn), Brian Ichter, Quan Vuong, Lachy Groom, and Adnan Esmail, drawn from Google, Stanford, UC Berkeley, and the technology industry. [10][11][14]

The company's stated aim is to bring general-purpose artificial intelligence into the physical world by building learning algorithms for a single model that can control any robot to do any task. [2] Hausman has framed the goal as a general [robot foundation model](/wiki/robot_foundation_model) rather than software tied to one machine. In one account of the company's plans he said, "What we're doing is not just a brain for any particular robot. It's a single generalist brain that can control any robot." [11] Co-founder [Sergey Levine](/wiki/sergey_levine) has summarized the ambition more bluntly: "Think of it like ChatGPT, but for robots." [18] The underlying premise is that the scaling behavior seen in language models might also apply to robot control if a model is trained on enough varied data across many robot bodies and environments. [1]

## What are the π0 and π0.5 models?

Physical Intelligence released its first model, π0, on October 31, 2024, describing it as a generalist robot policy. [6][12] π0 is a [vision-language-action model](/wiki/vision_language_action_model): it takes in camera images and a text instruction and outputs continuous motor actions for a robot. [6][12] The model is built on PaliGemma, a vision-language model from Google with about three billion parameters, and it adds an "action expert" component that produces smooth, high-frequency motor commands using a technique related to flow matching. [12] π0 was trained on data from several different robots performing dozens of tasks, combined with the [Open X-Embodiment](/wiki/open_x_embodiment) dataset, and it could be fine-tuned to new tasks with roughly one to twenty hours of additional data. [6][12] In demonstrations it folded laundry, cleared and bussed tables, and handled other manipulation tasks. [6][12]

In February 2025 the company open-sourced π0, releasing the weights and code in an experimental repository called [openpi](/wiki/openpi), together with checkpoints for platforms such as the [ALOHA](/wiki/aloha_robot) arms and the DROID setup based on Franka arms. [12] In April 2025 it published π0.5, a successor described as a vision-language-action model with open-world generalization. [1] π0.5 was trained by co-training on heterogeneous data, including data from multiple robots, high-level semantic prediction, and web data, and it was shown controlling a mobile robot to clean kitchens and bedrooms in homes that did not appear in its training data, carrying out multi-step jobs lasting several minutes. [1] The company later announced further models in the same family. [2]

| Model | Released | Type | Key point |
|-------|----------|------|-----------|
| π0 (pi-zero) | October 31, 2024 | Vision-language-action policy | First generalist policy; built on PaliGemma plus a flow-matching action expert [6][12] |
| π0-FAST | February 2025 | Autoregressive VLA | Uses the FAST action tokenizer; released alongside the open-source openpi code [12] |
| π0.5 | April 2025 | Vision-language-action model | Open-world generalization to unseen homes via heterogeneous co-training [1] |

## How much funding has Physical Intelligence raised?

Physical Intelligence raised an early seed round of about 70 million dollars at a valuation near 400 million dollars, with backers including Thrive Capital, Khosla Ventures, Lux Capital, OpenAI, and Sequoia Capital. [11][14] In November 2024 it raised about 400 million dollars at a valuation around 2 billion dollars, in a round associated with Jeff Bezos, Lux Capital, and Thrive Capital, with participation from OpenAI, Redpoint Ventures, and Bond. [11][14]

In November 2025 the company raised a Series B of about 600 million dollars at a valuation of about 5.6 billion dollars, led by CapitalG, the independent growth fund of Alphabet, alongside Lux Capital, with participation from Bond, Redpoint, Sequoia Capital, T. Rowe Price, and Jeff Bezos, bringing total funding to roughly 1.1 billion dollars. [15][16][17] In March 2026, Bloomberg reported that the company, then employing about 80 people, was in talks to raise about 1 billion dollars more at a valuation exceeding 11 billion dollars, with Founders Fund and Lightspeed Venture Partners said to be involved alongside returning backers Thrive Capital and Lux Capital. [18]

| Round | Date | Amount | Valuation | Notable investors |
|-------|------|--------|-----------|-------------------|
| Seed | 2024 | ~$70M | ~$400M | Thrive, Khosla, Lux, OpenAI, Sequoia [11][14] |
| Series A | Nov 2024 | ~$400M | ~$2B | Jeff Bezos, Lux, Thrive, OpenAI, Redpoint, Bond [11][14] |
| Series B | Nov 2025 | ~$600M | ~$5.6B | CapitalG (lead), Lux, Bond, Redpoint, Sequoia, T. Rowe Price, Bezos [15][16][17] |
| Reported talks | Mar 2026 | ~$1B (reported) | >$11B (reported) | Founders Fund, Lightspeed, Thrive, Lux (reported) [18] |

## Why does Karol Hausman's work matter?

The work places Hausman within a wider research direction in [embodied AI](/wiki/embodied_ai) that tries to apply the methods of large pretrained models to physical robots. [1][2] Supporters of this approach argue that a single foundation model, trained on diverse robot data, could let robots generalize to new tasks and settings instead of being programmed for one job at a time. [1][11] The approach remains unproven at scale, and the π0 and π0.5 reports note that the models still make frequent mistakes in both planning and motor control. [1]

Hausman's research has been covered by general news outlets, including reports on Google's robot learning work and on Physical Intelligence's models, and his writing has appeared in technical venues. [3][6] He has been featured in interviews about the company and the prospects for general-purpose robots, and he is listed as a co-founder of Physical Intelligence by several of the firm's investors. [1][10][11]

## Facts

| Field | Detail |
|-------|--------|
| Name | Karol Hausman |
| Occupation | Roboticist, entrepreneur |
| Known for | Co-founder and CEO of [Physical Intelligence](/wiki/physical_intelligence); robot learning research (SayCan, RT-1, [RT-2](/wiki/rt_2)) |
| Current role | Co-founder and CEO, Physical Intelligence (since 2024) |
| Past roles | Staff research scientist, [Google Brain](/wiki/google_brain) and [Google DeepMind](/wiki/google_deepmind); adjunct professor, [Stanford University](/wiki/stanford_university) |
| Education | B.Sc. and M.Sc., mechatronics, Warsaw University of Technology; M.Sc., robotics, Technical University of Munich; Ph.D., robotics, University of Southern California (2018) |
| Doctoral advisor | Gaurav Sukhatme |
| Fields | [Robotics](/wiki/robotics), [robot learning](/wiki/robot_learning), [reinforcement learning](/wiki/reinforcement_learning), [vision-language-action models](/wiki/vision_language_action_model) |
| Notable products | π0 (pi-zero), π0.5 robot foundation models |
| Company valuation | ~$5.6 billion (November 2025 round) [16][17] |

## References

1. Physical Intelligence. "π0.5: a Vision-Language-Action Model with Open-World Generalization." Physical Intelligence blog and arXiv:2504.16054, April 22, 2025. https://www.pi.website/blog/pi05
2. Physical Intelligence. "Physical Intelligence (π)." Company website, accessed 2026. https://www.pi.website/
3. Karol Hausman. "Karol Hausman." Personal website, accessed 2026. https://karolhausman.github.io/
4. Stanford University. "Karol Hausman." Stanford Online instructor profile, accessed 2026. https://online.stanford.edu/instructors/karol-hausman
5. University of Southern California. "Karol Hausman." Robotics and Autonomous Systems Center, Robotic Embedded Systems Laboratory, accessed 2026. https://robotics.usc.edu/resl/people/57/
6. Physical Intelligence. "π0: Our First Generalist Policy." Physical Intelligence blog, October 31, 2024. https://www.pi.website/blog/pi0
7. Ahn, M., Brohan, A., Hausman, K., et al. "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances." arXiv:2204.01691, 2022. https://arxiv.org/abs/2204.01691
8. Brohan, A., et al. "RT-1: Robotics Transformer for Real-World Control at Scale." arXiv:2212.06817, 2022. https://arxiv.org/abs/2212.06817
9. Google DeepMind. "RT-2: New model translates vision and language into action." Google DeepMind blog, 2023. https://deepmind.google/blog/rt-2-new-model-translates-vision-and-language-into-action/
10. Sequoia Capital. "Karol Hausman." Founder profile, accessed 2026. https://www.sequoiacap.com/founder/karol-hausman/
11. Glasner, J. "Robot Brain Startup Physical Intelligence Raises $400M At $2B Valuation." Crunchbase News, November 2024. https://news.crunchbase.com/ai/robot-brain-startup-unicorn-physical-intelligence-bezos/
12. Daniel, A. "Physical Intelligence Unveils Robotics Foundation Model Pi-Zero." InfoQ, December 2024. https://www.infoq.com/news/2024/12/pi-zero-robot/
13. ResearchGate. "Karol Hausman." Research profile, University of Southern California, accessed 2026. https://www.researchgate.net/profile/Karol-Hausman
14. Sacra. "Physical Intelligence valuation, funding & news." Sacra company profile, accessed 2026. https://sacra.com/c/physical-intelligence/
15. The Robot Report. "Physical Intelligence raises $600M to advance robot foundation models." The Robot Report, November 2025. https://www.therobotreport.com/physical-intelligence-raises-600m-advance-robot-foundation-models/
16. Bloomberg. "Robotics Startup Physical Intelligence Valued at $5.6 Billion in New Funding." Bloomberg, November 20, 2025. https://www.bloomberg.com/news/articles/2025-11-20/robotics-startup-physical-intelligence-valued-at-5-6-billion-in-new-funding
17. Axios. "VCs are funding the AI-powered robot revolution." Axios, November 21, 2025. https://www.axios.com/2025/11/21/robots-physical-intelligence-ai
18. TechCrunch. "Physical Intelligence is reportedly in talks to raise $1B, again." TechCrunch, March 27, 2026. https://techcrunch.com/2026/03/27/physical-intelligence-is-reportedly-in-talks-to-raise-1-billion-again/