Work / Covariant

RFM-1 launch

Company
Covariant
Year
2024
Role
Lead designer, in-house
Scope
Product launch, video

With RFM-1, a new Robotics Foundation Model, Covariant robots can communicate with people in natural language and better understand the physical world.

Traditionally, robots have been programmed to repeat the same action over and over. Since 2017, Covariant's AI has let robots understand their environment and decide how to act.

The marketing team worked closely with AI research and company leadership on a series of assets announcing the breakthrough to investors, customers and the wider AI community. We worked to a tight deadline so the launch could be showcased at Modex, an industry trade show.

Launch clip for social, featuring Covariant cofounder and chief scientist Pieter Abbeel.
It’s a clear sign that the artificial intelligence systems that drive online chatbots and image generators will also power machines in warehouses, on roadways and in homes.
Cade Metz, The New York Times
RFM-1 language-guided programming demo.
RFM-1 physics and video generation demo.
The startup has been building the robot equivalent of an LLM database.
Brian Heater, TechCrunch
How RFM-1 tokenizes many kinds of input to generate many kinds of output.
RFM-1’s AI generates realistic videos of robots picking.

In the news

  • The New York Times
  • TechCrunch
  • MIT Technology Review
  • Wired
  • The Wall Street Journal
  • Fast Company
  • The Verge
  • AIM Research

Credits

Art direction and design
Elisa Massenzio
Scripts
Anu Saha, Ally Lynch
Video production
Elisa Massenzio, Christian Svizzero
Animations
Elisa Massenzio
Social strategy
Lottie Smith