Is there something about consciousness that gives humans the ability to have creative breakthroughs, that can't be captured (or simulated) by AI? 

If there isn't, then AGI/ASI should "retire" us, eventually. 

If there is, what should we do?

For true sports fans, every game their team plays is a brand new episode of their favorite reality TV show.

New episodes multiple times a week, new cast members every season, a steady stream of drama.

Why isn’t there an all-access paid broadcasting experience for sports teams?

If we live in a simulation, I think it’s more likely to be running for entertainment than science.

The simulation hypothesis has gained serious traction over the past two decades. The logic is straightforward: as technology advances, simulated realities will eventually become indistinguishable from base reality, and since there can be countless simulations but only one base reality, the odds suggest we’re in a simulation.

But while most discussions about this focus on scientific motivations—I suspect a more straightforward purpose: entertainment.

If David Deutsch’s many-worlds interpretation of quantum physics is correct, quantum computers leverage computation power from parallel universes. A quantum computer with 1 million qubits would theoretically use 2^1 million parallel universes—an unimaginably large number.

If we live in a finite multiverse, what happens when we exceed its capacity?

The economics of foundational AI models may resemble those of the airline industry.

Most of the value created accrues to the manufacturers, the energy suppliers, and the world at large—while the entities actually running the flights capture very little.

How far are we from a world where you can watch—or rather, experience—Neuralink livestreams?

Imagine millions of people tuning in daily to watch Elon Musk’s daily livestream, not just seeing his life but experiencing his unfiltered thoughts in real time.

Of course, there is a massive caveat to this literal thought experiment—that consciousness is sufficiently materialistic to be captured and shared. But if that assumption holds, the implications are fascinating.

Someone ought to build a marketplace for human judgment (taste).

As AI continues to drive down the cost of creation, the real bottleneck in future knowledge work won’t be AI-inference tokens; it will be human-judgment tokens.

The ability to discern quality, to exercise taste, to make the right call—these will be the scarce resources that matter.

I think this business will look like some combination of Fiverr/Upwork and GLG, but I doubt it will be either of them.

The emergence of AI insurance may be a necessity, because using AI in the real world has more to do with governance than capability.

The key question isn’t whether an AI can replicate the work of a software engineer — it’s whether society will allow it to, without proper mechanisms for allocating risk and accountability.

In other words: how is the duty of care distributed in a world of human-AI labor?