Xi (Peter) Chen is an American computer scientist who is most notable as one of the co-founders of the AI robotics company Covariant. Chen, who currently serves as the company’s CEO, was an early employee at OpenAI and worked under AI identity Pieter Abbeel at the University of California, Berkeley.
Education and early career
Chen completed his Ph.D. in Computer Science at the University of California, Berkeley, under the tutelage of Pieter Abbeel.
At Berkeley, Chen’s interests revolved around reinforcement learning, generative models, and finding ways to equip machines with the capacity to understand and interact with complex environments.
In 2013, Chen launched the online auction platform Sellegit Inc. The company, which provided a safe, efficient, and fun way to buy and sell items, was created at Berkeley and was initially part of the university’s SkyDeck Accelerator Program.
In March 2016, OpenAI announced that Chen would be joining the company as a research intern over the summer. He was also joined in the same intake by Rocky Duan, a then-robotics Ph.D. student who would later found Covariant when Chen and two others.
Over this time, Chen co-authored several academic papers on topics such as generative adversarial nets (GANs), how to train them, and baseline estimation in policy gradient algorithms. Chen also devised a benchmark suite of tasks to quantify progress made in the domain of continuous control.
Dinner with Rocky Duan
Chen remembers the point at which he decided to leave OpenAI. He was at a small restaurant in Oakland with co-founder Rocky Duan and, over dinner, the pair discussed a paper on teaching robots to learn new skills or adapt to new scenarios quickly.
However, both acknowledged the existence of a gap between the abstract pursuits of academia and the immense body of industry knowledge possessed by companies on the customer side. On a practical and specific level, this translated to a need for autonomous robots over those that could only perform one task repeatedly.
Chen, Duan, Abbeel, and Tianhao Zhang started Covariant in 2017 to address the need for autonomy in warehousing and logistics.
While other AI start-ups tended to use as much off-the-shelf hardware as possible, Chen noted that because Covariant was founded by a quartet of researchers, the company took a different path: “Covariant started from a very different place. We started with pure software and pure AI. The first hires for the company were all AI researchers. We had no mechanical engineers, no one in robotics. That allowed us to go much deeper into AI than anyone else.”
The company’s Covariant Brain AI system was released soon after and, in recent years, has seen its uptake increase.
In an article announcing a Series C funding round of $75 million, for example, Chen explained that Covariant experienced 600% growth over 2022 with leading companies turning to AI robotics “to decrease labor costs, increase throughput, and control profitability.”
- Peter Chen is an American computer scientist who is most notable as one of the co-founders of the AI robotics company Covariant. Chen currently serves as the company’s CEO, was an early employee at OpenAI, and also earned his doctorate under AI identity Pieter Abbeel at Berkeley.
- In March 2016, OpenAI announced that Chen would be joining the company as a research intern over the summer. He was also joined in the same intake by Rocky Duan, a then-robotics Ph.D. student who would later found Covariant with Chen and two others.
- Chen and Luan discussed a need for automation in robotics over dinner one night, and started Covariant to develop warehouse and logistics robots that could learn and operate autonomously in unpredictable, real-world environments.
Connected Business Model Analyses
Stability AI Ecosystem