Generative AI and Moats

Quick musings on generative AI’s impact on traditional media incumbents and the opportunity for founders to rethink markets.

Successful startups are born from change. Regulatory updates, cultural movements, and new technologies introduce vulnerabilities for incumbents that fail to adapt.

Generative AI is one of those big changes.

Every day new, wildly impressive tech demos of AI-generated videos, songs, and memes go viral on Twitter. You too can create your own sexy werewolf romance novels covers with a few words of text (keep reading). And the results keep getting better, fast.

While it’s difficult to predict how things will play out, it’s clear that incumbents that invested years building content libraries will be challenged.

For many use cases, on-demand generative AI media will deliver better results than the incumbent's solutions. Unlike traditional static media, generative AI media can adapt in real-time to the user's requests, preferences, and context. It can appeal to the smallest niche. And it can do so without licensing restrictions or costly human labor, infinitely.

In the future…

A Spotify competitor could generate entirely new music that matches the user's query. It won't be limited to existing music catalogs, birthing truly personalized and novel sounds. Suno is an early example.

The next Pinterest could generate house renovation inspiration on the fly, adapted to one's home, budget, and style. Catbird is exploring this.

A future Wattpad could write fan fiction instantly. It might include the reader's friends or relatable experiences that deliver a more engaging story. Storybooks does this for children’s books.

Of course, incumbents can also leverage this technology. But if it interferes with their business model or existing stakeholders, they may struggle to adapt.

Artists and major labels won't be too pleased if Spotify shifts a large portion of its audience toward royalty-free music consumption.[1]

Ecommerce companies will be upset if discovery of their products collapse as as attention shifts toward synthetically generated images.[2]

Writers at Wattpad won't be happy if sales of their werewolf romance stories drop as readers seek AI-generated novels.[3]

It's a classic case of innovator's dilemma.

Other moats

The defensibility of content libraries may be at risk when synthetic content is just as good (or better) and available at a fraction of the cost. Of course, some libraries are at more risk than others. Spotify’s vast network of music will always be relevant and desired by listeners. I’ll never stop listening to Porter Robinson’s Worlds album. On the other hand, Pinterest’s database of images have less timeless, cultural relevance.

In a future where media is abundant and cheap (and humans are synthetic), I’m increasingly bullish on companies that establish a moat with real humans.

People will continue to desire human connection, professionally (a la LinkedIn) and socially (a la Instagram). Yes, the Replika's and Character.ai's of the world will also be successful in satisfying our primal desire for entertainment and play, but they’re not a complete replacement to the personal connection and opportunities we gain from other people. Furthermore, people are status-seeking monkeys and need other humans to see them flex to feel acknowledged.

And to be clear, Spotify, Pinterest, and Wattpad may thrive in this new world if they can adapt. But until then, there's an opportunity for new startups to reimagine (literally) media consumption.

[1] Spotify could empower artists and rights holders to opt in to the use of their likeness and sound in the generation of new music and in return, earn a cut of revenues.

[2] Pinterest has the ability to present merchant’s product within synthetically generated images, driving even more attention and higher conversion rates.

[3] Wattpad has an opportunity to help storytellers and IP holders create more content for their fans.