Recently, Eminence Ventures hosted a Future Tech Demo Day centered on "AI + Application Innovation," bringing together founders, researchers, and operators across technology, product, industry, and investment.

In AI's next phase, value will be defined not only by model capability, but by how effectively applications turn that capability into real-world adoption.

At the event, Yanpei Cao, Chief Scientist of VAST, and Hang Chu, Co-Founder & CEO of Viggle.ai, shared perspectives on 3D and world models and real-time, on-device interactive content. Joy Dai, Managing Director of Eminence Ventures, also offered a systematic look at the thinking behind our investments in Pixverse, Viggle, and VAST. For us, these are not isolated portfolio stories, but part of a broader multimodal thesis spanning content generation, real-time interaction, and spatial intelligence.

Over the past few years, AI has moved through rapid hype cycles, but multimodal AI has remained one of the few themes with lasting momentum. We began investing in this space as early as 2022 because it sits at a rare intersection: it benefits directly from advances in foundation models while also being one of the fastest areas to translate technical progress into visible, shareable, and monetizable product experiences.

At the same time, multimodal competition is not just about technology. It is equally about timing, execution, and the ability to align product, capital, and go-to-market strategy. In our view, the market is now shifting from model benchmarking to product delivery. As core model capabilities mature, the key question is no longer simply who has the strongest model, but who can bring those capabilities to users more efficiently, more reliably, and in ways that fit real workflows.

That is why we remain highly convicted on the application layer. Pixverse, Viggle, and VAST each represent a different point in the migration of value within multimodal AI: the restructuring of video workflows, the rise of real-time interactive interfaces, and the long-term infrastructure opportunity represented by 3D and world models.

From our perspective, the next wave of multimodal opportunities will belong to teams that can turn advances in technology into durable product value, user adoption, and commercial outcomes.

‍