AI found the perfect strategy... from 2013

Yesterday on the live I told the story about the time I asked AI how to get 100,000 email subscribers.

I did the thing all the gurus tell you to do when using AI for this stuff...

... Deep research mode
... Go read the whole internet
... Find the patterns with most potential
... Come back with a few recommendations

It chugged away for fifteen minutes and then announced it had found the perfect strategy, the exact one a marketing brand used to add 300,000 subscribers in a year.

Obviously I was pretty damn excited.

So, I asked it to lay it out for me. Know what that "breakthrough strategy" was?

... A lead magnet with Facebook ads.

Right. Anybody who's been marketing longer than a week has that one in a drawer somewhere. So I asked when the case study was published.

2013.

Fifteen minutes of deep research to hand me a thirteen year old playbook. A playbook that worked insanely well back then, but has since become table stakes.

A strategy that no longer works because free lead magnets have no real value when everyone has AI.

And Meta ads have become so expensive that to make a free lead magnet work, you need a huge cash reserve to carry the cost.

That's the whole problem in one story.

AI predicts the average of the internet, so the moment you hand it the thinking, you get the average answer back with total confidence.

It picked that example because so many people had referenced it, linked to it, and copied it, AI thought it was the best thing out there.

If I'd followed that plan in 2026, I would have lost time, money, and my sanity.

The live was about what to do instead, and the replay's up now. It's about 50 minutes.

... The 10/80/10 rule, which is what decides whether you get consistent output or the usual slop
... A live terminal run where I add a book to the site from scratch. Data pulled, editorial written, page built and published while everyone watched. Half a penny.
... Why the whole thing runs on Haiku, the old cheap model, and why paying for the clever one would actually make the output worse
... The 40x ROAS ad my meta system told me to put the whole budget behind. It had one pound of spend and a single lucky sale behind it, which is exactly why you stay in that last 10%.
... The full 5-system blueprint mapped to ACCER, the growth team running my book site for less than $5 a day

​→ Watch the replay here​

The site's at 14,471 books now. 125 of them went up on the day of the live. Seven platforms get posted to daily. I don't touch any of it.

If you want those same five systems in your business, that's Automated Growth.

Ten seats, and it's capped at ten because I'm doing the setup work with you, not selling you a course to go and watch on your own.

$1,997 one time. It goes to $3k after this first cohort, cause the founder price is what you get for going first while I sharpen the thing.

Enrolment closes Sunday at midnight EDT and we start Monday. 14-day refund, and you own the systems outright rather than leasing them off me.

If you'd rather check it's actually a fit before you decide anything, there's a link on the page to book a quick call with me.

​→ Have a look at what's inside Automated Growth​

Pete "AI strategy is shallow as a puddle" Boyle