Foundation models scaled with data and compute. Does the same playbook apply in the physical world?
Join us for a short panel during #CoRL2026 in Austin on Physical AI: what scales, what doesn’t, and what’s missing. We’ll dig into:
-
Whether scaling laws hold for physical AI the way they did for foundation models
-
Robot learning across environments and applications, from autonomous vehicles to dexterous manipulation
-
Closing the sim-to-real gap
-
Data quality: what “good” looks like, and where the bottlenecks are
A networking mixer with hors d'oeuvres and drinks follows the panel.
Panelists
-
Arthita Ghosh, Robotics & World Models, Sieve
-
Saurabh Nair, Applied Science, Wayve
-
Joshua Manela, Founding Engineer, Bedrock Robotics
-
Sai Charitha Akula, Senior Research Scientist @ Meta Reality Labs
-
Isha Mehrotra, Investor, Uncork Capital (moderator)
Space is limited. Please RSVP early to reserve your spot.
Agenda
SponsorsSieve is the data lab for real-world intelligence. Founded in 2022 and based in San Francisco, it builds multimodal training datasets and evaluations for frontier AI. Sieve works with leading AI labs, Fortune 100 companies, and fast-growing startups. It operates global data collection and exabyte-scale infrastructure, and treats data creation with the same rigor as model development.
Uncork Capital is a seed-stage venture firm that invests in founder potential long before it’s obvious. The firm was founded in 2004 as one of the first seed funds in Silicon Valley. Portfolio companies include Warp Sciences, Thinking Machines Lab, Tailscale, LaunchDarkly, GPTZero, Fitbit, and more. It invests across AI infra, developer tools, and frontier technology, including autonomy and robotics.
Full text on Luma
.