What AI Did Next
We analysed 13 of the most watched AI channels on YouTube so you don't have to. The findings are free — this is where the conversation starts.
3,500,000 Views
One video asked the questions nobody else was asking. The audience response was overwhelming. We found out why.
9 to 1
The ratio of optimistic to cautious signals across every video we monitored. That imbalance is the story.
2015 vs Now
The man now accelerating AI faster than anyone on earth once called it "probably the greatest threat to the continued existence of humanity." That distance deserves an explanation.
If you've spent any time recently trying to understand what AI actually means for your life, your work, or your future — you'll have noticed something.
Everyone is very excited. And somehow, despite all that excitement, you still don't have the answers you came looking for.
That's not a coincidence.
The AI content world is enormous, fast-moving, and overwhelmingly optimistic. Talented, genuinely enthusiastic people are producing extraordinary amounts of content about what AI can do, what tools to use, and how to make money from it. Most of it is useful. Some of it is remarkable.
But the harder questions — what does this actually cost, who bears the risk if it goes wrong, and why does the person building the most powerful AI on earth keep changing his story about whether it's going to save us or destroy us — those questions are almost entirely absent from the conversation most people are most likely to encounter.
We decided to find out why.
We built an engine. We collected data. We read the transcripts, counted the mentions, mapped the sentiment, and looked for the gaps. And what we found was both more straightforward and more unsettling than we expected.
This report is what the data told us. No hype. No agenda. No breathless enthusiasm for tools we haven't tested or claims we can't verify.
Just honest intelligence — for anyone in the world who wants to think carefully about what is actually happening.
When we collected this snapshot in April 2026, Claude was already the most mentioned tool across our entire dataset — 1,518 mentions across 11 channels, far ahead of ChatGPT. The practitioner world had moved before the mainstream noticed. That's exactly the kind of signal an intelligence engine is built to surface.
We found workflows described as costing "$0 per month" that quietly excluded existing subscriptions and API costs running to thousands of dollars. The fine print matters.
And it got ten times more views than anything else we monitored. The audience for honest AI content is enormous. Almost nobody is serving it.
That wasn't a critic writing that. That was the company itself, in its own internal documentation. We found it so you didn't have to.
$500 billion committed. Datacentres potentially obsolete within three years. A historical pattern that has repeated across railways, telecoms, and the internet boom. The question isn't whether AI is real. It's who bears the cost if the financial architecture around it becomes unstable.
We believe the people who need honest AI intelligence most are the ones least likely to pay for it before they've seen it. So we're not asking you to.
Enter your first name and email below. We'll send you the full report immediately. You'll also be the first to know what the engine finds next.
That's it. You can unsubscribe any time. We won't make that difficult.
No spam. No selling your data. Unsubscribe any time. We mean that.