newsroom
Enterprise AI Adoption · Media/entertainment workflows are the beachhead for generative video adoption (Fox case study)
now playing · Enterprise AI Adoption
Vertical integration (proprietary models + GPU clusters) is the defensible moat for creative AI
Krea's journey from API wrapper to training its own diffusion model (Crea 2) on owned GPU clusters in Ohio demonstrates that controlling the full stack — model architecture, post-training d…
Post-training data curation, not base model access, is the new differentiator in open-source AI
With base architectures (Transformers, diffusion) commoditized and open-source models (Flux, Llama) closing the gap to closed-source, competitive advantage shifts to post-training: human-cu…
a16z platform value-add is real for AI infrastructure: GPU access, enterprise G2M, recruiting
Beyond capital, a16z's operating teams deliver concrete execution leverage: GPU procurement via Nvidia/network relationships, enterprise intros to media/tech buyers, and recruiting pipeline…
Generative AI pricing will converge to utility model; current subsidy wars are unsustainable
Application-layer companies are subsidizing inference costs (up to 75%) to acquire users, but this mirrors early telecom/data pricing that eventually converged to flat-rate subscriptions. T…
Media/entertainment workflows are the beachhead for generative video adoption (Fox case study)
Fox Corporation uses Krea for agentic video workflows (replacing production shoots with generative clips), but most users still prefer human-in-the-loop: generate scenes, edit in After Effe…
Innovation in US, optimization in China, regulation in Europe — three-speed AI world
The founders observe a structural split: fundamental model innovation concentrates in US (SF), efficient deployment/optimization in China, while EU regulation (AI Act, GDPR) creates complia…