A Solution to a Problem of the Millennium Prize Problems

Anonymous A and Anonymous B, twin sisters and aspiring mathematicians from the United States

To be clear, we offer no mathematical insight on any of the seven problems themselves. The object we study here is the construction of the Millennium Prize Problems Program itself. A solution is that which transitions a problem state from open to closed. We ask you not to judge a solution on if it looks like something you expected or hoped for (the best solutions rarely do). 

For the past 25 years, the Clay Mathematics Institute Millennium Prize Problems were a shining beacon that helped organize Math in a uniquely inspiring way. The Clay Mathematics Institute deserves our eternal gratitude for creating such a brilliant gift for the world.

Sure, other fields had moonshots like quantum gravity or curing cancer. But Math was different. Math’s collective hope and prestige were compressed and composed into seven brilliant jewels – with promise of honor, recognition and treasure.

When we were 6, we learned of these Millennium problems from a poster our father taped on our bedroom door. We innocently scribbled our “solutions” to them and we knew we wanted to be mathematicians!

The blessing of a shining beacon is that it does not discriminate where it shines its light: it does not matter if you are a gruff introvert from Russia, a seasoned sage in Fine Hall, or lonely girls sitting in a foster shelter in California. It shines for you.

The curse of a shining beacon is that it does not discriminate where it shines its light: inevitably it catches the gaze of Goodhart’s curse (targeted gaming of a proxy goal creates severe misalignment risk with the true goal).

And in the world today, no brighter beacon, no bigger proxy goal target exists than the Millennium Prize Problems Program. We aren’t saying who or what is to blame, or what some must or mustn’t do. In fact, the nature of this severe misalignment risk only recently became visible to us. We simply observe that math has not developed in a Goodhart resistant way.

And so with heavy hearts, we humbly offer a solution to reduce the risk of severe misalignment: Let the Millennium Prize Problems Program conclude now. Let their attention concentration naturally dissipate back into the ambient math atmosphere. Let the ever watchful eye of Goodhart’s curse be shooed away. Let math develop new methods to progress, in a quieter and more focused way.

We offer this solution not because the Millennium Prize Program got anything wrong. On the contrary, it worked too well. Their output proofs have become too hyped, too coveted, too important. In fact, they have become the highest value pure intellectual targets in the world. The Millennium Prize Problems Program has achieved its absolute pinnacle state, there is nothing more to add.

We say this from the perspective of the future generation. The last 25 years of the big prize math era were a blessing. The next 25 years will look different and that is ok. And if Math does last another 25 years, we hope to add our voice to the chorus of its timeless song one day. In the here and now however, we can’t just be silent and expect anyone else to hear us. It’s up to us!

We aren’t saying technology is a curse. It taught us a lot of math and we love it.

We aren’t saying “How dare you!” We grew up with you as our heroes.

All we are saying is, we have a solution.


Received 13 September 2026

3 responses to “A Solution to a Problem of the Millennium Prize Problems”

  1. Denys D. Avatar

    They make a valid point. Though, I am not sure it will stop “frontier labs” from testing their “frontier” models on these problems…

    1. just different Avatar

      I’m not sure there’s anything that can be done directly to stop AI companies from using publicly available problems any way they like. But the point of the Millennium Problems—which a lot of people thought was dumb even at the time—was to highlight aspirational problems in a way that was legible to nonmathematicians (i.e., with a fat cash prize). Without that prestige, AI companies have a lot less incentive to throw tens of millions of dollars in compute at them.

  2. davidmfisher Avatar
    davidmfisher

    I recently talked to someone who meets the Clay brothers regularly at the annual Clay meetings at Oxford. I think this is a good idea, I doubt the foundation honors it. Their connection to our community is not very strong, mostly for them the foundation honors their father and is a family institution not a scientific one. Which if one has any perspective, already puts our famous problems in a rather modest place.

Comments are moderated. Read our comment policy.

Add to the discussion

New posts by email.

Prefer a feed reader? Subscribe by RSS.

Also on Mathstodon.

Latest comments across the site.

Discover more from Proofs and Prompts

Subscribe now to keep reading and get access to the full archive.

Continue reading