Physical Intelligence is building pi-zero into a generalist robot policy
An October 31, 2024 pi-zero release puts Physical Intelligence around a robot policy trained across images, text, and actions.

Pi-Zero Generalist Policy
Physical Intelligence published pi-zero on October 31, 2024 as its first generalist robot policy. The model spans images, text, and actions, then outputs low-level motor commands.
The pi-zero release says the policy can control different robots and can be prompted or fine-tuned for harder task scenarios. That places Physical Intelligence inside the research-lineage arc: a model breakthrough first, field proof later.
Robot Task Demonstration Boundary
The release uses folding a shirt and cleaning a table as examples of physical tasks. Those examples are model demonstrations, not audited customer deployments.
Pi-zero still has to transfer policy behavior from demonstrations to reliable work in customer environments. Task success rate by deployment site is the metric that would turn pi-zero from research artifact into operating system under Physical Intelligence.
Physical Intelligence Verification Boundary
Physical Intelligence has disclosed the October 31, 2024 pi-zero release, images-text-actions model scope, low-level motor-command output, multi-robot control claim, and prompted or fine-tuned task modes. Public material does not show customer deployment count, task success rate by site, intervention frequency, robot-body transfer cost, or field uptime. It also does not disclose pricing, customer retention, paid deployment count, or renewal rate. The next useful metric is verified task success rate per robot body by customer under Physical Intelligence.
Have a robotics update Korthos should review? Send news, deployments, product releases, funding rounds, research, or media to tips@korthos.xyz or reach out on X at @agkorthos.
- Physical IntelligenceCompany
Track the machine economy
Regular Korthos briefings on robotics, drones, physical AI, supply chains, funding, product launches, and the companies shaping the stack.