AI images are getting scary good (here's what we're doing with them)


For two years I've been watching AI image tools improve.

Every few months a new one would launch.
Every creator would hype it up.
Then I'd test it on real client work and the output was... fine.
Usable.

But not totally something I'm stoked about.

Recently that changed.

GPT Image 2 is good! Really good!!!

The thumbnails it's producing for the summit videos in our AI system is crazy.

All done with a training image and a prompt we've refined over 3 months.

So what we're doing now is going back through every video we've ever uploaded that still has potential and split testing new thumbnails.

But now we have a new problem...

The thumbnails are no longer the problem.

The intros are.

People click... watch the first few seconds... and bounce. Because the opener doesn't match the energy of what made them click in the first place.

So now we're rebuilding a new intro framework for every summit going forward.

This is what business actually looks like.

It's finding what works.
Making it 1% better.

Then noticing the next weakest link.
Fixing that.
Then the next one.

Over and over.

A few people have asked what I'm using to make all this happen.

Inside our automations... GPT Image 2 is doing the heavy lifting via Fal.

But when I'm making something one-off and want full creative control, I still use Magnific (used to be called Freepik). It's the cleanest tool I've found when you want to actually direct the output instead of just prompting and hoping.

There's 50% off the Business plan in our deals platform. Saves you up to $352 over the year. Worth grabbing if you make visual content regularly.

And speaking of deals... this weekend is probably the right time to grab AI Content Lab if you've been thinking about it.

AI Content Lab contains access to the deals platform for the next 3 months.

You can get $50 off with code MAYMADNESS. (ends Monday night)

I don't really use these emails to sell often (prefer to keep it educational and mostly as a way for me to gather my thoughts).

We have 4 emails coming out this weekend from the BigMoves AI brand about it all...

But the cliffnotes are:

  • we just added 6 epic content bots trained on the summit sessions
  • our members have saved $300k+ from the deals platform in the last 90 days
  • all AI Content Lab members get early access to the next summit sessions (AI Content Creator Summit) and those sessions will also help improve the existing AI tools we've released
  • (all existing AI Content Lab members get these updates for free)
  • the price goes up when the next summit launches

If you want that, cool. If not, no worries.

The summit will still be free and you can sign up in June when it goes live!

Talk soon, Mitch

Mitch Asser

Founder of BigMovesai & MakeBigMoves.ai - helping businesses implement AI effectively. This newsletter documents the journey.

Read more from Mitch Asser

Getting sick last year was one of the best things that happened to me. It didn't feel that way at the time. I had to let go of 3 team members while trying to get BigMoves AI off the ground. It was rough. But looking back... it forced something I wouldn't have done otherwise. It forced me back into the weeds of business. Here's the trap most experienced business owners fall into with AI. If you've been running a team for years, you've been trained to delegate. Something not worth your time?...

I was in Singapore recently and caught up with a mate who coaches some of Australia's most successful entrepreneurs. He's written books on it. Spent years figuring out what separates people who consistently perform from those who just work hard and stay stuck. He told me my edge is 'the innovator.' Someone who has a feel for how things work across all types of businesses and niches... and can piece together old methods with new technology. I didn't disagree. But I also hadn't thought about it...

One of the most challenging decisions business owners have today is: When do I go all in on AI? And the answer isn't straight forward. For example, we're continuing to debate whether to rebuild our entire website and move off Wordpress to a custom solution. Long-term it makes total sense. Every few months it becomes easier and easier. And I've seen a number of people spend 3-4 weeks moving to an AI first, vibe-coded solution for serious websites. (of course you can stand up a few simple pages...