I've had a front-row seat to the AI ecosystem for a while now, because at this point pretty much every major consumer AI company that wants into the North American or European market crosses my desk eventually. That vantage point led me to a conclusion I want to walk through in this one: we've finally hit the precipice of end to end AI content. Not text, that got cracked years ago, but realistic, engaging video that can build brands, creators, and entire businesses on autopilot.
I break down the three upgrades that just converged to make this possible: near-perfect AI lip sync, image and video realism that's basically indistinguishable from the real thing, and the back-end agent infrastructure that lets people spin up content at a scale that wasn't feasible even a few months ago. I get into the tools driving each piece (Veo, Seedance, Kling, ElevenLabs, MCP, agents running on your desktop) and what it actually means when an AI agent can script, voice, generate, title, thumbnail, post, and respond to comments with no human in the loop.
Then I get practical about the money. I lay out the three biggest opportunities I'm seeing right now: directly monetizing your own AI content (and why AI/business niche CPMs make this work without huge scale), selling content as a service to companies looking to slash their social budgets, and the fast-growing world of fully AI-generated ads. If you work in content, this is the landscape you're operating in today. Huge thanks to AnySearch for sponsoring. It's the search infrastructure built for the AI era, giving your agents one API key into specialized data sources (SEC filings, GitHub, legal, research, security feeds) that Google and Perplexity just don't surface. Link in the description.
CHAPTERS: 0:00 The State of AI Content Creation Right Now 0:40 Why End to End AI Video Is Finally Here 1:06 What This Video Covers 1:24 Sponsor: AnySearch Search API for AI Agents 3:24 Stage 1: AI Text and Scriptwriting 3:55 Stage 2: AI Visuals and Voice (Veo, Seedance, Kling, ElevenLabs) 5:03 The Missing Pieces: Lip Sync, Realism, and Workflows 5:41 AI Lip Sync Is Nearly Perfect 6:21 AI Image and Video Realism Is Game Over 6:47 The Back End: Automating Content With AI Agents 7:41 Agents That Run an Entire Content Business 8:32 Why Speed and Volume Beat Quality For Now 8:57 Human Gated Social Platforms Are Coming 9:23 Opportunity 1: Monetize AI Content Directly 9:59 Opportunity 2: Sell AI Content as a Service 10:37 Opportunity 3: Fully AI Generated Ads 11:00 Recap: Three Ways to Win With AI Content ------------------- Clips channel @jonlawclips
I do NOT sell any products of my own. Watch out for scams.
So I've gone on to watch the AI ecosystem for content and tools for creators really evolve
and grow over the past few years. And that's because at this point just about every single
major consumer AI company that wants to access the North American or European market is going
to flow through me at some point. And so I get this very interesting perspective where I see
what the big companies are doing and kind of their tools or angles and how they want to promote
their tools and services and products. And then also the startups and some of the more creative
and kind of esoteric ideas out there. And my conclusion in terms of AI for content and for
social media is that just in the past quarter, the past two or three months maybe, have we
gotten to a point where I think we're on the precipice of end-to-end AI content. And by
that I mean video content, not text. That was crack years ago. But AI being able to end-to-end
create video that looks realistic, that is engaging, and being able to build brands, build
creators, build businesses out there through social media and content distributors and platforms. In
this video I want to talk through that change, what's important now, and how I think you can go
and build businesses like that. And then not in this video but on this channel we cover all of
the different tools that materially enable you to do that. So that's going to be this video. Let's
hear a word from the sponsor and then we'll get into it. So guys, your AI agents are searching the
wrong internet. The most valuable data on the internet lives outside of what Google and
Perplexity actually search. You're talking SAC filings, GitHub repos, legal databases, research,
security threat feeds, all the data that professionals and agents actually need. And
yet all of those popular search tools, your Perplexity, your Brave, your Google, and even
most of the agents you're already running miss all of it. And that is exactly the gap that any
search is built to close. And any search is the search infrastructure built for the AI era. One
API key gives you instant access to a huge database of specialized data sources across finance, law,
science, security, geography, and more. And the routing is fully automatic. So you don't tell
any search what database to query. Its intent classifier reads what you're asking, sends it to
the right sources, and then when multiple sources hit, it fuses the results, ranks them by quality,
deduplicates, and finally hands back clean markdown that your LLM can actually parse. So here's what
that actually looks like in practice. Here's me researching a product for an upcoming video
and trying to find actual user feedback past a wall of affiliate listicles. So here's same query
with two tools side by side. So I didn't tell it where to look. It read the query, routed it to
the right places, and then handed back something that's actually usable. Now again, this is not
designed for you to read. It is designed for your agents to use. So you plug it into your cloud
desktop or windsurf as an MCP server and then run your agent like you normally would. And then when
it needs to look something up, it can route automatically through any search. And because
it's agent native, the responses are low token. So your costs are very reasonable on longer run.
Now any search is freely available for developers to try, and it can be integrated into any AI
agentic workflow. So the link is the first one in the description, and the GitHub is linked there
as well. So go give it a try. Again, that's the link on the screen or in the description. And now
back to the video. So text again was cracked years ago. That was the first stage in being able to
automate a lot of content because a lot of video content does involve words. I'm speaking to you
right now. And even if I'm on camera, maybe I'm reading from a script that was completely written
by AI, right? That could have been the case. That is the case for a lot of people out there
as of 2023, 24, 25. So that was really stage one. And that absolutely was a jump, but it's not
anywhere close to being able to end to end run these things with AI. And then you saw the
development of two things simultaneously, which is AI being able to create visuals on one side
of things and then voice on the other side. And those still aren't all the way there. That was
like 24, 25. So kind of the past two and a half years where those tools really started to hit
their stride, where you have the image and video tools getting a lot better. That's your VO, your
Cdance, your Kling. And then AI voices now have gotten to a point where they're extremely realistic.
They have tonality. I think the easiest example of that is going to the ChatGBT app and just
talking to the voice agent and just talking to the voice. It is really good. And then you see
these other platforms like Noise, 11Labs that have really, really good voices. And so you combine
those things and you can now automate a lot of videos where maybe there's no human in them,
but there are visuals or there's something being said, or it uses footage from somewhere else.
And then you have the voice or the narration on top where AI is able to do the scripting
and the voiceover. But those two pieces are not the end game. That was the start of the end game.
But what hadn't been cracked were realistic humans and then really good lip sync. And then finally,
the ability of AI to manage workflows across different apps and integrations and basically
be the infrastructure or the behind the scenes for this kind of thing. And that is what in the
past quarter we have finally hit where those three verticals for upgrades have converged
and are starting to hit their stride. They're not perfect, but it's all come together. And now it's
just a realism. It's kind of a touch up game as far as, oh, the lip sync is 98% accurate or the
human realism is 3% off, but we are pretty much there. And that's the first time I can confidently
say that after looking at the last three, four or five years. And so let's break that down more
specifically. Number one, we have lip sync. There are tools out there now, which are just about
perfect with this. Like if you took a real human voice and the AI generates the video or vice versa,
you took an AI voice lip syncing to a real human or AI to AI, it is perfect. Like you cannot tell
the difference. And then maybe in the past three months, you've seen videos where it's like a human
and then an AI kind of filter where it replicates the motion of the human, but with an AI human,
those are perfectly realistic. Tons of people are creating content with them. And the realism
on pictures nowadays, especially is again, perfect. You can go create videos and a lot of people do
this, especially with models and very attractive people to sell stuff where it is, you, you just,
you cannot tell the difference. It's game over in terms of realism. And those things will only
continue to get better at an accelerating pace. So those were two of the main pieces, which are
more visual in nature, and you'll hear news about them. The third, which we talk a lot about on this
channel that is less reported out there is just the backend. Like how would you, for example,
go create a thousand videos a day with AI across a hundred different media channels, right? That's
kind of the organizational or the infrastructure question, which maybe you and I won't do that.
And maybe that's not the ideal strategy for actually deriving value or a business from
these tools, but that is a million percent what a lot of people are going to do. And now you start
to have the tools to do it. You have the MCP infrastructure, you have, for example, Cloud and
Chrome, you have agents that can run on your desktop and your computer. They can have their
own social media accounts. They can operate 24 seven autonomously. That whole layer combined
with these other visual tools mean that you can have agents that directs these other sub tools,
and they can autonomously do everything from script to generate voices, to generate visuals,
to titling, to thumbnails and covers, to then posting, responding to comments. They can do
it all. They can do everything involved in a modern content business, whether a solo creator
posting on social media, more of a responding to news or kind of coverage type content, all the way
through to creative animations, whatever it is, AI can now do all of it. Is the quality a hundred
percent there? No, but that's the curve we're seeing now where it can all be done and people
are doing it. They will start to do it in bulk, although I think we're still really early. And
then any kind of a quality spread there between people doing it manually and the AI doing it,
right? The AI content and AI content businesses is going to compress and compress until it gets
to zero. We're not there yet, but we are close. And what it makes up today in the quality,
not being there, it makes up for a thousand times over in speed and automation and volume,
which is why people are already starting to pick these tools up and run these businesses
in this kind of way. So is that good? You know, in a lot of ways, I think not. And I think what
we're going to start to see is just human gated social platforms where people will algorithmically
try to differentiate human versus AI content and just kind of put a wall up between the two. But
for now it's just free game. And a lot of people are going to try to scale this thing up as far
as they can and have tens, hundreds, thousands of different social accounts or businesses that
are autonomously run with agents. And in terms of how I would frame the different opportunities
there, if that's something you're thinking about, number one is just directly monetize that content,
right? Like being monetized on YouTube or X. I think Facebook now pays creators up to like
three grand a month just for posting reels because apparently for whatever it is, option one is just
directly monetize that content that works, you just need a lot of scale. And it's very niche
dependent. But you can do really well that way, right? YouTube can pay people like the AI and
business niche 10 2030 50 plus dollar CPM. So you don't need that much scale to make a pretty good
income doing it that way. What a lot of other people are starting to doing or the second kind
of big category there in terms of the opportunity is just selling those things to companies. What
you're doing there is going into a company that spends 50 grand a month on their social team and
making content. And you just say, I will rip out this set of deliverables that you spend 30 grand
of that 50 grand a month on, and I'll do it for 10. The company is happy, they say 20 grand,
you're happy because it only costs you five. And if the deliverables are the same, because the
quality is there, that's a win, win, win, except for the people that got fired, which is very not
good for them. But maybe they go do other things inside the company. Let's let's hope that works
out. And then the third big kind of category or opportunity I'm seeing there is just ads where a
lot of platforms and companies and individuals are switching their ad funnel and their content for
ads over to fully end to end AI generated ads. So if you are going to do well, generate value out
there with these kinds of advancements. Again, as a recap, it's probably going to be making your own
content and monetizing that directly, selling the kind of content you've made to other companies
as a service or for use in ads for maybe your existing product or service business. So that
is what I wanted to cover in this one. For better or worse, we certainly live in interesting times,
and there is a lot of opportunity out there. And I think that is an area worth exploring
if you work in content. That's the video. Thanks for being here. And I'll see you in the next one.
Cheers.