Why “Bubble” Language Fails Bubbles pop when people realize the asset doesn’t deliver the promised value. But AGI doesn’t need to deliver the promised value (artificial god). It just needs to deliver enough cognitive replacement to offset the missing workers. That’s a much lower bar. And we’re already seeing narrow AI hit that bar in specific domains. So even if we get “disappointing AGI” - something that’s 10x better than GPT-4 but nowhere near sci-fi singularity - that might still be enough to justify current infrastructure investment. The bubble critics are assuming the payoff needs to match the hype. But the payoff just needs to match the problem. And the problem is massive.

The Bubble That Has To Happen Traditional bubble: Capital floods into an asset class beyond any reasonable return expectation. This situation: Capital floods into the only visible solution to structural labor shortage. Yes, most AI companies will fail. Yes, current valuations look absurd by normal metrics. But if the alternative is watching developed economies shrink 20-30% over the next two decades, then even a 10% chance of AGI working justifies trillion-dollar bets. It’s not irrational exuberance. It’s rational desperation.

AI Isn’t a Bubble - It’s a Bailout Everyone calling AI a bubble is using the wrong reference class. They’re comparing it to dot-com, crypto, tulips - speculative manias disconnected from fundamentals. But what if the fundamental is “civilization as currently structured cannot function without replacing the missing workers”? Then suddenly the “insane” capital allocation makes perfect sense. You’re not pricing AI companies on 2025 revenues. You’re pricing them on “what percentage of the $100+ trillion global economy can we save from demographic collapse?”

Reading Zeihan’s *The End of the World is Just The Beginning* He clearly understands fiat’s structural problems, but then frames both GME and Bitcoin as examples of fiat *failure* rather than what they actually are. GME wasn’t “gamer mania”, it was market structure meeting coordinated information flow. Bitcoin’s success isn’t a symptom of fiat failure, it’s an architectural response to it. Conflating monetary dysfunction with market mechanics and treating a new monetary protocol as speculative mania wasn’t something I saw coming. #bookstr

#photography

#photography

Be like water

Claude Code on the web has been interesting. I think Claude Code coming to mobile is the highlight, though.

I enhanced the #Plurcast --nostr-pow (#NIP13) implementation by using a custom mining setup rather than the library's .pow method, so now it's multithreaded and finishes quicker.

I've been revising my #GitHub Readme, sharing more about my project visions, and I added my Bitcoin/Lightning addresses.

I'm excited to get out and do some #photography today

#art

Bech32 encoding: note1 prefix = Nostr event ID q character in bech32 = zero/low value in the underlying data Multiple qs at the start = leading zeros = proof of work

proof of work forces intentionality. Mining a #Nostr note means you had time to cancel the thought broadcast.

thinking about cryptographic attestation chains

#Plurcast now supports --nostr-pow, enabling #NIP13 to embed proof of work into your Nostr notes. The previous note had a --nostr-pow 22 and took maybe 10-15 seconds to mine.

NIP-13 is a Nostr Implementation Possibility (NIP) that defines a method for generating and interpreting Proof of Work (PoW) for nostr notes. PoW is used as a bearer proof that relays and clients can universally validate with simple code. It helps to deter spam by requiring computational effort to be proven on a note. The difficulty is measured by the number of leading zero bits in the note ID.

I watched The Big Short last night, it was really good. Definitely catered towards people that enjoy finance, but they also made an attempt at bringing everyone along, and there were a few moments that were pretty funny.

#art