Google turns the NYC subway into a Veo submission funnel

What the campaign actually asks people to do

The "Imagine If…" project, announced November 14, is a joint effort by Google and OUTFRONT that spans all five New York City boroughs for four weeks. The flow is deliberately simple: a commuter spots the campaign on OUTFRONT's digital screens in the transit system, scans a QR code to submit an idea, and instantly receives a shareable visualization of their imagined version of the city.

The prompts Google offers as examples are concrete and playful — Manhattan skyscrapers as vertical farms, Brooklyn brownstones "shimmering with neon magic." That framing tells you the intended input: short, visual, place-specific ideas that map cleanly onto what a text-to-video and image model can render.

Two Google products are named explicitly: Veo for video and Nano Banana for image generation. Both are also pitched at the end of the post as available in the Gemini app for anyone not in New York, making the campaign a funnel toward the consumer product as much as a local art event.

Artists as the filter between the public and the model

The most telling design choice is that the public does not use Veo directly. Google recruited five local artists — Ariana Cimino (Staten Island), Jeff Wave (Queens), Lauren Camara (Bronx), Molly Goldfarb (Manhattan), and Subway Doodle (Brooklyn) — described as the campaign's "visual architects." They select written submissions from their own borough and then use Veo to turn those into finished video artworks.

That structure inserts human curation and authorship between raw public prompts and the outputs displayed city-wide. The instant visualization a commuter sees from the QR code is separate from the polished artist-made pieces that appear across MTA screens and, ultimately, in Times Square. It is a two-tier system: immediate machine output for the individual, curated artist output for the public gallery.

The artist roster leans toward established, distinctive styles — Camara's cut-paper portraits, Goldfarb's acrylic and marker work, Subway Doodle's Emmy-winning blue monsters. Google is borrowing their credibility and aesthetic control rather than presenting generative video as an unmediated tool.

The physical distribution plan

The campaign runs on thousands of digital screens across MTA subway stations, described as turning "New York's transit network into a moving gallery of collective imagination." The four-week run culminates December 14 in Times Square, where the most compelling artist pieces are shown on OUTFRONT's screens at the north end of the square.

OUTFRONT's role matters: the out-of-home advertising network provides both the intake surface (the screens with QR codes) and the display surface (station screens and Times Square). Google supplies the models and the artist partnerships. The result is a closed loop where the same screen infrastructure collects prompts and shows results.

What staging Veo as street infrastructure signals

The specific implication of "Imagine If…" is that Google is treating generative video not as a lab capability to demo but as something to embed in physical public space with human intermediaries. Rather than pointing people to a chatbot, it places Veo behind a QR code in the subway and behind five named artists who own the final output.

For teams building with generative models, the notable pattern is the separation of instant self-serve output from curated, human-authored output. Google gets viral participation from the QR flow while protecting the quality and attribution of what actually goes on public display through the artist layer. The Gemini app pitch at the end confirms the campaign's dual purpose: a New York cultural event and a distribution channel for Veo and Nano Banana.