The Ellisons won Paramount, and then Warner. Now comes the hard part: taking on Netflix and YouTube.

It took Larry and David Ellison two years to assemble a media behemoth.

Analyst Rich Greenfield at Lightshed Partners thinks he knows the game plan the Ellisons will have to pursue: Not just bulking up his existing Paramount and HBO services by buying high-profile programming like the UFC and hiring the guys behind "Stranger Things," but opening up his streamers to stuff outsiders make. That's the core of YouTube's success story, and increasingly seems to be the direction Netflix is headed.

And it seems all the more urgent to figure this out soon, before a wave of AI-generated video changes the very idea of traditional media library: If anyone with a computer can spit out a movie or a TV show, what's the value of the catalogs media companies have spent years and fortunes assembling?

Peter Kafka: David Ellison bought Warner to turn Paramount into a company that can compete with Netflix. What does he think he's building?

One of these big decisions that's sitting out there is how do they actually combine all of these assets, especially on the streaming side? They've said they want to have one service. Is it all one? Do they raise the price to accommodate all of the HBO content? Do they keep the HBO brand? There are a lot of things that they're working on, and it's really hard to analyze right now because they haven't said very much since they announced the acquisition.

Add BI in Google so our reporting is easier to find when you’re searching for what matters.

You really like Taylor Sheridan. I think most people watch Taylor Sheridan content or kids' content like SpongeBob [on Paramount]. But if everyone is watching the same thing, it's very hard to train an algorithm. What you really need is many different signals. That's what makes YouTube so good at discovery. That's what makes Netflix better than everyone else.

I think the real question is, are you combining these companies and aggressively investing in substantially more content? Because both of these streaming services are undernourished and need a lot more content to win engagement and time spent per day.

If you talk to many of the AI video experts that are out there, they'll say generating a minute of video in the lab nearly instantaneously already exists.

We're basically there. The important point is not when is the technology there, but when is it not cost-prohibitive to roll this out in mass scale. My guess is it's within the next 12 to 18 months. Maybe I'm naive, maybe it's even sooner, but it's clearly coming.

YouTube is a platform for other people's content, wherever it's made anywhere in the world. And I think Netflix is moving in that direction. Do they create a lot of content on their platform? They do, but if you look at the total amount of content, it's still other people's content by and large on the platform. And as they move more into podcast and shorter-form video content, they are becoming and flexing this platform focus.

If you're trying to win, you're going to have to be a platform. I don't think any of these companies are going to be able to create enough content on their own.