Welcome to the AI For Humans newsletter!

This week, the Arizona Cardinals tried to use AI to make a fun NFL schedule release video.

Reader, they did not succeed.

The video got publicly trashed by rival NFL teams, Pat McAfee, and a wall of fans who absolutely turned on it within hours.

As of this writing, the original post has fewer than 1200 likes. The LA Chargers' very-not-AI Halo-themed video, for comparison, hit 100,000.

Even worse, because the Cardinals used AI to make it, anti-AI sentiment became a dominant point of the criticism. Other teams even went out of their way to mention their videos were made by human hands.

But here's the thing, and this is what I want to dig into today:

The Cardinals video wasn't bad because it was AI.

It was bad because it was bad.

That distinction is going to matter a LOT more in the months ahead.

Let's get into it!

Please support AI For Humans by learning about our sponsors below:

Ready to transform artificial intelligence from a buzzword into your personal revenue generator

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

Get Your Guide

If you've been watching public sentiment around AI, none of this is surprising.

The polls keep showing AI getting worse numbers from the general public, not better. "AI slop" has become a phrase regular humans now say out loud. Eric Schmidt, former Google CEO, was just booed every time he mentioned AI at a commencement speech.

Even Seth Rogen got in on it this weekend, making the case that AI-generated content is creatively bankrupt:

But you might think sports fans would be different.

Sports fans love spectacle, special effects, fast cuts, weird stunts. If anyone was going to be open to AI-generated content, it should be them.

And this is one of the NFL’s biggest off-season content weeks of the year.

The NFL has turned schedule release day, yes, the schedule, the literal list of who plays who, into one of the biggest social media moments of the entire offseason.

The Seahawks' fragrance commercial parody this year was a masterpiece, complete with Marshawn Lynch unveiling a perfume called "F--k You." The Chargers did Halo. The Colts redid the entire intro to The Simpsons. Most of these go viral and travel online for days.

So when the Cardinals decided to go AI… they were doing so in a much larger way and with a much larger audience than any piece of AI content has had in a long time.

And they massively failed.

But, more importantly, they didn’t get the idea right in the first place.

Again, it wasn’t that the fact that the Cardinals used AI to make the video, it’s that they did it so poorly.

Last week, I wrote about Markos Slavnic's viral AI pigeon video, also AI-generated, also reached a massive audience, but was fully celebrated… by people who normally hate AI video.

The tools didn't matter. The idea was clever, the writing was sharp, the craft was visible.

The Cardinals video was the opposite of all that.

The central concept? A Zoom call. Between NFL mascots. Voiced by a "Director of Mascot Matchups" whose audio fades in and out for three minutes.

What is this…2020?

As someone who had to produce a TV show in the Covid era, I can tell you the Zoom comedy angle has been properly milked for everything it has.

But the part that tipped it into the very bad was something they did at the end of the video that (I think) they felt was a cool, clever peek behind the curtains.

They lauded the fact that, besides the Cardinal mascot, all the other mascots on the call were AI-generated.

On the AI side, sure it’s cool that we can do video-to-video matching but as an NFL franchise there’s more going on here.

You're showing your audience that you didn't even bother to film with humans inside your own video. You couldn't even be bothered to schedule with the other mascots.

Even worse…

It’s an idea that AI actually wasn’t needed for at all.

It signals exactly the things you don't want to signal during a brand-building moment:

Laziness. Cheapness. Lack of Care.

(The fact that the Cardinals are also one of the worst-run franchises in pro football probably doesn't help, but I'll stay in my lane.)

The Packers, who absolutely did shoot their video with a handmade set and physical models, dunked publicly:

That one's going to sting for a while.

Love this newsletter? Forward it to one curious friend. They can join in one click.

Here's the thing I keep coming back to as AI tools get better and the slop gets sloppier:

The public is not anti-AI video. Not really.

The public is anti-bad.

"Bad" has always been a problem. Cheap-looking ads. Phoned-in content. Things that feel like they were made by a committee that didn't care.

These have always gotten dunked on. AI just gives audiences a new label to slap on the same old feeling.

When something AI-made is genuinely good, Markos's pigeons, Theoretically Media's short films, the AI-assisted commercials you didn't realize were AI-assisted, most people don't actually mind.

In fact, they actively tend to not really care.

What people mind is when a billion-dollar football franchise tries to swap out human creativity and labor for a bad idea executed poorly.

Or worse, when that NFL franchise signals they’re forward thinking and oh so cutting edge.

The lesson here is the lesson it's always been:

Care more than the audience expects you to. Make it look like a person actually thought about it. If you're going to use AI, use it because it gets you to a better idea, not because it lets you skip the work of having one.

The audience is paying attention. They were paying attention before AI, and they're paying triple attention now.

And, if you screw it up, they’re going to let you know about it.

-Gavin

This week on AI For Humans: Google I/O is coming!👇

3D modeling is one of those skills that's always been brutally hard for non-experts to get into. Try opening Blender and the interface alone is intimidating.

But recently, I pointed my Claude Code at Blender, and the results have been kind of shocking. Claude can just do stuff in there that would have taken me hours of YouTube tutorials to figure out manually.

Now X user Neilsonks has taken this to the next level and built an open-source 3D modeling pipeline that takes a single image and walks it all the way to a fully interactive 3D environment, fully driven by Claude Code.

I haven't actually tried this one yet (sorry, the weekend got away from me), but the demos look genuinely wild. If you've ever wanted to build a 3D world without learning Blender from scratch, this is probably the most accessible on-ramp we've ever had.

Let me know what you make.