A look at Meta's recent hires for its nascent superintelligence unit, including Pei Sun, who developed two generations of Waymo's perception models
Meta's new hires offer a glimpse into its nascent superintelligence unit, aimed at making the social media company's AI capabilities more competitive with industry leaders.
Context & Ripple Effects
Meta’s superintelligence effort was being assembled through targeted recruiting across frontier-AI specialties. Pei Sun’s work on two generations of Waymo perception models gives the unit a visible connection to AI systems that interpret the physical world.
The hire fits a broader buildout: Meta also recruited researchers from OpenAI’s Zurich office and Google DeepMind for the effort, before later adding Apple AI talent and creating an applied AI engineering organization to support it.
First-order effects
- Meta adds Pei Sun’s perception-model experience to its nascent superintelligence unit, broadening the technical background represented in the team.
- The unit becomes a more concrete competitive vehicle for Meta’s push to improve its AI capabilities, rather than a purely stated ambition.
Second-order effects
- Repeated recruiting from leading AI groups, including researchers from OpenAI’s Zurich office and a Google DeepMind principal researcher, raises retention and hiring pressure on the frontier-AI organizations supplying that talent.
- Meta’s ability to combine specialist recruits with its existing product reach becomes a more important differentiator as it tries to translate research hiring into competitive AI capabilities.
Third-order effects
- If this pattern continues, frontier-AI competition will be shaped not only by model releases but by whether large platforms can institutionalize cross-domain research teams and the engineering organizations around them.
- The later move to form a dedicated applied engineering group suggests that elite-research recruiting increasingly requires a parallel path for deployment, integration and organizational execution.
The trend: Frontier AI is becoming an organizational arms race in which large platforms recruit scarce specialists and build dedicated structures to turn research talent into product capability.