We created an 'arm farm' where robots learn to grasp objects by trial and error, improving their grasping abilities through pooled sensor data.

Automat is developing a more effective robotic process automation solution that eliminates the need for costly consultants, leveraging recent AI advancements.

Data collection for robotics is challenging; we can't scrape the necessary data from the internet, and we rely on expensive teleoperation methods to teach robots.

Lilac Labs is automating the drive-thru order-taking process, addressing inefficiencies caused by outsourcing these jobs to low-wage countries.

We are at an exciting moment where advancements in AI algorithms are finally enabling robotics to realize its potential.

In our projects, we aim to create products that are computationally, robotically, and materially novel, transitioning from traditional assembly to using robots and microbes.

The rise of AI coding agents is akin to the impact of the combine harvester on agriculture, where fewer farmers are needed to produce more food, suggesting a similar trend in software engineering.

As planners, we should first address bottlenecks preventing progress, such as the lack of humanoid robots or automated biology labs.

Robotics is progressing rapidly, and with the integration of AI, we can expect robots to communicate and interact with humans more effectively.