
Excited to see @Tubi featured in @WSJ following great conversations with @bellelin_, with perspectives from both me and @anjsud, on how we’re using AI to win over Gen Z audiences by improving all three sides of our marketplace (Viewers, Content Creators, Advertisers).
A few key ideas behind our approach:
1. Hyper-personalization at scale: combining deep learning + LLMs, including Tubi Foundation Model (built on open source pretrained model and fine tuned on our proprietary data) to understand not just what you watch, but why
2. AI is expanding discovery: as people turn to conversational interfaces and AI agents, we’re meeting viewers at the moment of intent (including with our recent native ChatGPT app launch)
3. One AI foundation, multiple unlocks: the same content understanding that powers discovery also drives more relevant, scene-aware advertising such as our our Tubi Moments ads product
4. Creator economics are changing: AI is a tool for creators, lowering the barrier to storytelling and expanding who gets to create
5. Built for experimentation and velocity: Tubi is fueled by experimentation (almost 2000 per year), and AI is rapidly accelerating that pace across the entire company
In this era, the most important attribute is learning fast, constantly reimagining how things can be built using the new tools, and staying scrappy (we’re even willing to rack NVIDIA GPUs in the office if it can help unlock more experiments for our eng team!).
Big shoutout to @santacruzseana for helping bring this story to life.
Link: wsj.com/cio-journal/wh…
English























