The Future of Content

The Future of Content

The Future of Content

The Future of Content

The Future of Content

The Future of Content

The Future of Content

The Future of Content

The future of content is not what you think

The best part about making the AI Faceswap video is that I ended up talking with Riley Brown for almost two hours on Friday.

We dove down a really interesting rabbit hole about the future of content.

The AI Faceswap technology is currently v1, and it’s already incredibly convincing. It’s going to get much better.

We predict, in about 12 months, anyone in the world with a reasonably high powered computer, is going to be able to perfectly render any face on any source video with a single click.

AI voice technology will follow a similar path.

Not only will you be able to turn yourself into existing people (e.g., I become Ryan Gosling), but you’ll also be able to digitally construct new characters that don’t currently exist.

We’re already seeing this with platforms like HeyGen and Synthesia…artificially created, life-like characters.

And of course these characters will be optimally attractive, optimally personable, optimally vulnerable. Max 100 on all stats like Madden “create character” mode.

Okay, so what does that mean for content creators?

What it means is that the moat around content will no longer be technical ability (to edit or film), physical attractiveness, or vocal delivery.

So what’s left?

My first thought was the “storytelling and script writing.” You might be able to artificially manufacture the “who is saying” but the “what is said” is still an artform, right?

Not exactly.

There are AI models being trained on every YouTube video ever produced. Every Tiktok video with over 1M views. Every great video ever.

Today, ChatGPT is shitty at writing scripts for short-form video. This is because, as Riley pointed out to me, it wasn’t trained on video content transcripts. It was trained on things like blogs and wikipedia.

Once the YouTube-specific AI models are trained on every great YouTube video ever made, the “storytelling and script writing” will also be replicated by AI better and faster than any human could produce it.

So where does that leave us?

It leaves us in a place where anyone in the world can produce any amount of content, at optimal quality, with the optimal story, with perfect looking characters that have perfect voices and are increasingly optimized based on all meaningful viewer retention metrics.

….Yeah, I know.

And maybe it takes longer than 12 months…let’s say 36 months to be conservative. In 3 years, this will for sure happen.

So what are the possible scenarios here?

  • The government could intervene and stop AI platforms from enabling this type of god-mode generation. They would do this, effectively, in the name of “saving art.” If this happens, then perhaps there would still be a moat in content around the creator storytelling skillset

  • The government could allow the technology to be created, but mandate some sort of profit share with the creators whose content trained the models. Wouldn’t it make sense that if MKBHD had 200 of his videos scraped to train the super YouTube AI, he’d deserve some portion of the profits from the product that enables anyone in the world to make better videos than him? I’d think so. So in this scenario, the creators with the largest back catalogues to train the models might profit handsomely

There’s a lot of ways you could play the “what if” game from here.

My heart wants me to believe that art will always exist and taste will always matter.

But sadly, I don’t think those are protectable either.

AI will be trained on every tastemaker ever. Meaning newly generated, fresh tastes will be the only thing that is “unique.”

But the AI will also adapt at increasing rates, so any novel “taste” that gains popularity will be quickly replicated and broadly distributed by AI.

Much more to dive into here and will be sure to have Riley on the podcast as an early episode when it launches.

— — — — — — — — — — — — —

If you enjoyed this post and want more like it, you should subscribe to me weekly creator journal, Blueprint. Each week, I share metrics, ideas, frameworks, and experiments designed to supercharge your thinking about content & brand building in the modern age.

The future of content is not what you think

The best part about making the AI Faceswap video is that I ended up talking with Riley Brown for almost two hours on Friday.

We dove down a really interesting rabbit hole about the future of content.

The AI Faceswap technology is currently v1, and it’s already incredibly convincing. It’s going to get much better.

We predict, in about 12 months, anyone in the world with a reasonably high powered computer, is going to be able to perfectly render any face on any source video with a single click.

AI voice technology will follow a similar path.

Not only will you be able to turn yourself into existing people (e.g., I become Ryan Gosling), but you’ll also be able to digitally construct new characters that don’t currently exist.

We’re already seeing this with platforms like HeyGen and Synthesia…artificially created, life-like characters.

And of course these characters will be optimally attractive, optimally personable, optimally vulnerable. Max 100 on all stats like Madden “create character” mode.

Okay, so what does that mean for content creators?

What it means is that the moat around content will no longer be technical ability (to edit or film), physical attractiveness, or vocal delivery.

So what’s left?

My first thought was the “storytelling and script writing.” You might be able to artificially manufacture the “who is saying” but the “what is said” is still an artform, right?

Not exactly.

There are AI models being trained on every YouTube video ever produced. Every Tiktok video with over 1M views. Every great video ever.

Today, ChatGPT is shitty at writing scripts for short-form video. This is because, as Riley pointed out to me, it wasn’t trained on video content transcripts. It was trained on things like blogs and wikipedia.

Once the YouTube-specific AI models are trained on every great YouTube video ever made, the “storytelling and script writing” will also be replicated by AI better and faster than any human could produce it.

So where does that leave us?

It leaves us in a place where anyone in the world can produce any amount of content, at optimal quality, with the optimal story, with perfect looking characters that have perfect voices and are increasingly optimized based on all meaningful viewer retention metrics.

….Yeah, I know.

And maybe it takes longer than 12 months…let’s say 36 months to be conservative. In 3 years, this will for sure happen.

So what are the possible scenarios here?

  • The government could intervene and stop AI platforms from enabling this type of god-mode generation. They would do this, effectively, in the name of “saving art.” If this happens, then perhaps there would still be a moat in content around the creator storytelling skillset

  • The government could allow the technology to be created, but mandate some sort of profit share with the creators whose content trained the models. Wouldn’t it make sense that if MKBHD had 200 of his videos scraped to train the super YouTube AI, he’d deserve some portion of the profits from the product that enables anyone in the world to make better videos than him? I’d think so. So in this scenario, the creators with the largest back catalogues to train the models might profit handsomely

There’s a lot of ways you could play the “what if” game from here.

My heart wants me to believe that art will always exist and taste will always matter.

But sadly, I don’t think those are protectable either.

AI will be trained on every tastemaker ever. Meaning newly generated, fresh tastes will be the only thing that is “unique.”

But the AI will also adapt at increasing rates, so any novel “taste” that gains popularity will be quickly replicated and broadly distributed by AI.

Much more to dive into here and will be sure to have Riley on the podcast as an early episode when it launches.

— — — — — — — — — — — — —

If you enjoyed this post and want more like it, you should subscribe to me weekly creator journal, Blueprint. Each week, I share metrics, ideas, frameworks, and experiments designed to supercharge your thinking about content & brand building in the modern age.

The future of content is not what you think

The best part about making the AI Faceswap video is that I ended up talking with Riley Brown for almost two hours on Friday.

We dove down a really interesting rabbit hole about the future of content.

The AI Faceswap technology is currently v1, and it’s already incredibly convincing. It’s going to get much better.

We predict, in about 12 months, anyone in the world with a reasonably high powered computer, is going to be able to perfectly render any face on any source video with a single click.

AI voice technology will follow a similar path.

Not only will you be able to turn yourself into existing people (e.g., I become Ryan Gosling), but you’ll also be able to digitally construct new characters that don’t currently exist.

We’re already seeing this with platforms like HeyGen and Synthesia…artificially created, life-like characters.

And of course these characters will be optimally attractive, optimally personable, optimally vulnerable. Max 100 on all stats like Madden “create character” mode.

Okay, so what does that mean for content creators?

What it means is that the moat around content will no longer be technical ability (to edit or film), physical attractiveness, or vocal delivery.

So what’s left?

My first thought was the “storytelling and script writing.” You might be able to artificially manufacture the “who is saying” but the “what is said” is still an artform, right?

Not exactly.

There are AI models being trained on every YouTube video ever produced. Every Tiktok video with over 1M views. Every great video ever.

Today, ChatGPT is shitty at writing scripts for short-form video. This is because, as Riley pointed out to me, it wasn’t trained on video content transcripts. It was trained on things like blogs and wikipedia.

Once the YouTube-specific AI models are trained on every great YouTube video ever made, the “storytelling and script writing” will also be replicated by AI better and faster than any human could produce it.

So where does that leave us?

It leaves us in a place where anyone in the world can produce any amount of content, at optimal quality, with the optimal story, with perfect looking characters that have perfect voices and are increasingly optimized based on all meaningful viewer retention metrics.

….Yeah, I know.

And maybe it takes longer than 12 months…let’s say 36 months to be conservative. In 3 years, this will for sure happen.

So what are the possible scenarios here?

  • The government could intervene and stop AI platforms from enabling this type of god-mode generation. They would do this, effectively, in the name of “saving art.” If this happens, then perhaps there would still be a moat in content around the creator storytelling skillset

  • The government could allow the technology to be created, but mandate some sort of profit share with the creators whose content trained the models. Wouldn’t it make sense that if MKBHD had 200 of his videos scraped to train the super YouTube AI, he’d deserve some portion of the profits from the product that enables anyone in the world to make better videos than him? I’d think so. So in this scenario, the creators with the largest back catalogues to train the models might profit handsomely

There’s a lot of ways you could play the “what if” game from here.

My heart wants me to believe that art will always exist and taste will always matter.

But sadly, I don’t think those are protectable either.

AI will be trained on every tastemaker ever. Meaning newly generated, fresh tastes will be the only thing that is “unique.”

But the AI will also adapt at increasing rates, so any novel “taste” that gains popularity will be quickly replicated and broadly distributed by AI.

Much more to dive into here and will be sure to have Riley on the podcast as an early episode when it launches.

— — — — — — — — — — — — —

If you enjoyed this post and want more like it, you should subscribe to me weekly creator journal, Blueprint. Each week, I share metrics, ideas, frameworks, and experiments designed to supercharge your thinking about content & brand building in the modern age.

The future of content is not what you think

The best part about making the AI Faceswap video is that I ended up talking with Riley Brown for almost two hours on Friday.

We dove down a really interesting rabbit hole about the future of content.

The AI Faceswap technology is currently v1, and it’s already incredibly convincing. It’s going to get much better.

We predict, in about 12 months, anyone in the world with a reasonably high powered computer, is going to be able to perfectly render any face on any source video with a single click.

AI voice technology will follow a similar path.

Not only will you be able to turn yourself into existing people (e.g., I become Ryan Gosling), but you’ll also be able to digitally construct new characters that don’t currently exist.

We’re already seeing this with platforms like HeyGen and Synthesia…artificially created, life-like characters.

And of course these characters will be optimally attractive, optimally personable, optimally vulnerable. Max 100 on all stats like Madden “create character” mode.

Okay, so what does that mean for content creators?

What it means is that the moat around content will no longer be technical ability (to edit or film), physical attractiveness, or vocal delivery.

So what’s left?

My first thought was the “storytelling and script writing.” You might be able to artificially manufacture the “who is saying” but the “what is said” is still an artform, right?

Not exactly.

There are AI models being trained on every YouTube video ever produced. Every Tiktok video with over 1M views. Every great video ever.

Today, ChatGPT is shitty at writing scripts for short-form video. This is because, as Riley pointed out to me, it wasn’t trained on video content transcripts. It was trained on things like blogs and wikipedia.

Once the YouTube-specific AI models are trained on every great YouTube video ever made, the “storytelling and script writing” will also be replicated by AI better and faster than any human could produce it.

So where does that leave us?

It leaves us in a place where anyone in the world can produce any amount of content, at optimal quality, with the optimal story, with perfect looking characters that have perfect voices and are increasingly optimized based on all meaningful viewer retention metrics.

….Yeah, I know.

And maybe it takes longer than 12 months…let’s say 36 months to be conservative. In 3 years, this will for sure happen.

So what are the possible scenarios here?

  • The government could intervene and stop AI platforms from enabling this type of god-mode generation. They would do this, effectively, in the name of “saving art.” If this happens, then perhaps there would still be a moat in content around the creator storytelling skillset

  • The government could allow the technology to be created, but mandate some sort of profit share with the creators whose content trained the models. Wouldn’t it make sense that if MKBHD had 200 of his videos scraped to train the super YouTube AI, he’d deserve some portion of the profits from the product that enables anyone in the world to make better videos than him? I’d think so. So in this scenario, the creators with the largest back catalogues to train the models might profit handsomely

There’s a lot of ways you could play the “what if” game from here.

My heart wants me to believe that art will always exist and taste will always matter.

But sadly, I don’t think those are protectable either.

AI will be trained on every tastemaker ever. Meaning newly generated, fresh tastes will be the only thing that is “unique.”

But the AI will also adapt at increasing rates, so any novel “taste” that gains popularity will be quickly replicated and broadly distributed by AI.

Much more to dive into here and will be sure to have Riley on the podcast as an early episode when it launches.

— — — — — — — — — — — — —

If you enjoyed this post and want more like it, you should subscribe to me weekly creator journal, Blueprint. Each week, I share metrics, ideas, frameworks, and experiments designed to supercharge your thinking about content & brand building in the modern age.

The Future of Content

The Future of Content

The Future of Content

The Future of Content

© WavyLabs. All rights reserved.