Miju Labs

← All buyers · Runway

A

Anastasis Germanidis

Co-founder & Co-CEO · Runway

high confidenceTier A · Talk to firstImage, video & audio modelsvideo · multimodal

Germanidis co-founded Runway and is now co-CEO alongside Cristóbal Valenzuela, having previously been CTO. He is the intellectual driver of Runway's pivot from video generation to general world models, and as of September 2026 he is publicly arguing that world models are the RL environments for everything, including robot policy evaluation. That pivot is the key fact for Julian: Runway's centre of gravity is moving away from filmmaker taste and toward physical-world simulation.

Background

Germanidis's early work (2015–2018) was in creative applications of machine learning — interactive film installations, GAN experiments, video-analysis tools — before co-founding Runway with Valenzuela and Alejandro Matamala Ortiz out of NYU ITP. Runway shipped Gen-1 through Gen-4 and landed on TIME's 100 Best Inventions of 2024 for Gen-3 Alpha.

By 2026 Runway is valued at around $5.3B and has repositioned around world models, robotics and real-time video, with a dedicated Runway Robotics effort and a research claim that simulated and real-world outcomes correlated at 0.95 across eight robot policies. His essays 'Towards Universal Simulation' (Oct 2024) and 'Real-World Superintelligence' (Jul 2026) set out the thesis.

What they run now

Runway's research direction: general world models (GWM Worlds 2, with what he calls a general action space merging navigation, manipulation and human interaction), real-time video, Runway Robotics, and a developer platform. He explicitly frames large-scale video pretraining as the advantage over domain-specific logs — a model trained on general video already knows how a pedestrian stumbles — and frames world models as evaluation substrate for policies, not just content generation.

What they buy

Inference. Runway's buying is shifting: less 'is this shot beautiful' and more 'does this simulation behave correctly'. That points toward physical-plausibility annotation, action-conditioned video data, robot-policy evaluation, and world-consistency rubrics. Creative taste data still matters for the film and enterprise business (Runway Studios, brand work) but it is no longer where the research money is going. No data vendor is named publicly.

Career

  1. 2018 – presentCo-founder & Co-CEO, Runway — Previously CTO; co-CEO with Cristóbal Valenzuela by 2026
  2. c. 2016 – 2018Researcher / creative technologist, NYU ITP
  3. c. 2016–2018New York University (ITP), Graduate work, interactive telecommunications (unconfirmed)

On the record

world models as RL environments · 2026

Says world models will become the RL environments for essentially everything. source

video pretraining beats logs · 2026-09

Argues broad video pretraining covers long-tail events that a driving fleet would log only a handful of times. source

simulation as evaluation · 2026-09

Cites 0.95 correlation between simulated and real outcomes across eight robot policies as evidence world models can replace physical evaluation. source

Interviews & talks

Why it matters here

Moderate, and worth being honest about. Runway was the obvious creative-taste buyer two years ago; in 2026 its research budget is chasing physical AI. Julian can still sell here, but the wedge is either (a) the enterprise/creative side under Jamie Umpherson and Anthony Maggio, or (b) reframing his offer as physical-plausibility and world-consistency judgement, which is still human judgement on video and which Runway's own eval claims depend on. Germanidis sets the frame; he does not sign small data contracts.

How to reach

Speculation: do not pitch creative taste to Germanidis. If approaching him, engage on evaluation validity — the 0.95 correlation claim invites the question of who grades the real-world outcomes, and expert human adjudication of sim-versus-real discrepancies is a defensible vendor product. He is active on X (@agermanidis) and responds to technical argument. The Runway AI Summit is a real access point.

What we could not establish
  • Whether Runway still buys creative-preference data now that research has moved to world models.
  • Who owns Runway's data procurement — no named role found.
  • His exact education and dates.

Why they are on the map

Promoted to Co-CEO Feb 2026; long-time research/technical head of Runway's models. source — evidence dated 2026-02

Decides whether the problem is funded.

How Runway buys data

Gen-4.5 (Dec 2025) cited 'significant advances in post-training' and ranked via blind pairwise human preference on the Artificial Analysis arena. 2026 hiring includes 'Research Engineer, Data Foundations' (design multimodal datasets for world models), 'MTS, Data Infrastructure' (billions of rows of multimodal data), and a 'Sr/Staff PM, ML Research'. Past content licensing deals (Getty Images 2023, Lionsgate 2024) show a legal/licensing buying path. No named human-data vendor found.

Co-CEOs (Germanidis, Valenzuela) since Feb 2026, and a CTO (Sindi). Research is led by Research Director Jimei Yang with research leads under him; there are separate Data Foundations and Data Infrastructure groups, plus Runway Studios (in-house creatives) and a Chief Creative Officer.

Known vendors: lmarena (Artificial Analysis arena for public ranking; not a vendor)

Sources

  1. Runway AI Summit — San Francisco 2026 (speaker list)
  2. Ground Truth vol. 4 — Anastasis Germanidis (2026-09-18)
  3. Runway Research — Introducing General World Models
  4. Anastasis Germanidis — personal site