My 2025 Predictions Scorecard

It’s that time of year again. Last December I published 12 predictions for 2025. Now it’s time to grade them: 5 right, 5 wrong, 2 still playing out. I batted .500—or .416 if you count the inconclusives as misses.

For comparison, my 2024 scorecard was 4 right, 5 partially right, 5 wrong.

This piece is free. My 2026 predictions will be partially paywalled. Subscribe now to get all of them.

Key

• ✅= Nailed it.

• ❌= Wrong

• ⏳= TBD / still playing out

✅ 1. There will be no AI-generated hits, blockbuster or niche

I can’t name one. Can you?

✅ 2. AI emerges as a “storytelling” performer

“AI will be less a storyteller and more a producer of faster, cheaper “performances” of stories.”

The release of OpenAI’s Sora 2 was proof in the pudding of AI’s ability to be an independent storyteller.

❌ 3. A messy debate about the “Quality” vs. Possibility of storytelling emerges

There was not much of a debate but rather a logical evolution of creator understanding. The volume of AI “slop”—low-value, high-volume videos—grew as the number of Hollywood productions declined. For example, nearly ten percent of the fastest-growing YouTube channels were using AI “slop” to grow.

My own sense from conversations over the past year and from November’s Artist & The Machine summit is that Hollywood studios, showrunners and screenwriters see the path (back) to quality through the possibility.

One point I made was directionally right: “Newfound efficiencies and possibilities in production models from AI will matter less if consumers increasingly prefer their favorite IP in different formats than TV or movies.” However, videos which emerged from Sora, Google Veo 3 and an orchestra of other platforms suggested consumers prefer likenesses—which is a form of IP—to their favorite IP—like Disney’s “Star Wars” or Warner Bros. “Harry Potter”.

⏳ 4. “The new business of living room content” redefines the role of producer

YouTube’s dominance of living rooms grew in 2025—it was up 2.1% year-over-year in Nielsen’s most recent “The Gauge”. This *is* happening. You can infer it from the decline of Hollywood productions and the growth of YouTube (above).

There are over 3 million creators in the YouTube Partner Program.

✅ 5. The value of IP libraries for TV series and movies continues to decline

This point may seem debatable given Netflix’s $82.7 billion price tag for Warner Bros. studios and streaming.

However, Disney took some tough feedback from consumers on the long-term value of its IP with box office disappointments “Captain America: Brave New World”, “Tron Ares”, “Thunderbolts*” and “Snow White”. Six Marvel shows failed to chart in Nielsen’s streaming ratings on Disney+:

• “Daredevil: Born Again”

• “Ironheart”

• “What If…? (Season 3)”

• “Your Friendly Neighborhood Spider-Man”

• “Eyes of Wakanda”

• “X-Men ’97”

Last, Disney implicitly conceded this point in its licensing agreement of over 200 Disney characters to OpenAI’s Sora. Effectively, the walls of the walled garden need to be loosened for brands to adapt to the generative AI era.

❌ 6. Netflix continues to pivot and evolve away from its legacy media movie and TV DNA

Its likely acquisition of Warner Bros. Discovery is a pivot towards that DNA.

❌ 7. A small business using Meta’s Movie Gen tools will create most successful AI-generated ad campaign of 2025

There were more controversial ad campaigns than successful ones.

Also, Meta’s AI efforts pivoted a few times, firing and hiring along the way. Chief scientist of Meta AI Yann LeCunn left with a skeptical prediction on LLMs.

Ironically, Meta made a terrific holiday ad without AI.