About that $1M math problem...

Read the Asterisk

Yesterday I told you OpenAI claimed to solve a Millennium Prize problem.

And I told you to read it as "OpenAI says."

Well, the fine print just landed.

Here's what actually happened, because it's a masterclass in how to read AI announcements.

The problem is Navier-Stokes.

The equations that describe how fluids move. Weather, blood flow, air over a wing, all of it.

For 26 years, nobody could prove whether the equations can "blow up," meaning spiral into infinite values from smooth starting conditions.

That's the million-dollar question.

OpenAI's clarification says their result proves blowup CAN happen...

Under a smooth external force.

That word "force" is the whole asterisk.

The actual Clay prize asks about the equations with NO external force.

So OpenAI now says, in their own words, the result does not claim the prize.

They also credited two human mathematicians doing closely related work at the same time.

So did the hype machine lie?

Not exactly.

The result is genuinely remarkable.

They deployed 10,000 AI agents on an unreleased model and cracked a variant of a 26-year-old problem in 88 hours.

The model found a vortex that tightens and spins ever faster while the fluid's energy stays bounded.

That's real mathematics. Real progress. Real proof.

And they say they've made substantial progress on ANOTHER Millennium problem since.

But "solved a Millennium Prize" and "proved a forced variant of one" are different sentences.

The internet only heard the first one.

Here's the lesson, and it's bigger than math.

Every AI announcement now comes in two layers.

The headline that spreads.

And the fine print that's true.

The gap between them is where you get fooled into rebuilding your stack, chasing the wrong tool, or betting on a capability that has an asterisk you didn't read.

So build the habit.

When something sounds impossible, wait 48 hours for the clarification.

It always comes.

The people who reacted to yesterday's headline are embarrassed today.

The people who waited for the fine print are exactly where they were, and better informed.

Be the second group. 🫡

Did You Know?

The Navier-Stokes equations were written down in the 1840s...

They run every weather forecast, every airplane design, and every model of blood flowing through your heart.

We use them billions of times a day.

And after 180 years, we still can't fully prove they always behave.

We built the modern world on math we don't completely understand. 🌊

🗞️ The Tip AI News 🗞️

The people building AI keep walking out the door.

Two more researchers just left their jobs at Anthropic and Google DeepMind.

Both went public about why.

Both said the technology may soon spiral beyond control, with the possible extinction of humans on the table.

Not activists. Not politicians.

Researchers who were inside the labs last month.

This isn't an isolated event, and that's what makes it worth your attention.

Over the summer, 1,100 employees across the major labs signed a letter asking Washington for a way to slow things down.

Last week, an Anthropic researcher resigned citing extinction risk.

This week, a new OpenAI board member said the industry is "not on track" to bring loss-of-control risk to an acceptable level.

And now two more walk out with the same warning.

The pattern is the story.

Now, here's how I think about it, because I refuse to give you either panic or dismissal.

These are serious people with front-row seats, and they're scared.

That deserves respect, not eye-rolls.

At the same time, the technology isn't going back in the bottle, and you still have a business to run.

So what does a sane person do with this?

Two things.

You stay honest about the risk instead of pretending the insiders are crazy.

And you keep building responsibly anyway, with the sandboxes, the scopes, and the human oversight this newsletter has been preaching all summer.

Those two things aren't in conflict.

They're the same posture.

Take the technology seriously enough to use it well, and seriously enough to respect what it could become. 🫡

😂Meme of the Day

Over to You...

Go wait 48 hours before you believe the next impossible headline...

And go build carefully this weekend. 🔥

Jeff J. Hunter 

Founder, AI Persona Method | TheTip.ai

Get paid to solve problems like this → AI Certified Consultant

“AI Money Group” to Learn to Make Money with AI