106 - Writer's Block

As I begin book #5 in my marathon through 10 new books, "Antifragile" by Nassim Nicholas Taleb, after already enjoying #2 on that list, his book "The Black Swan: Second Edition", I have to take a moment to give some unexpected praise to Yuval Noah Harari's book "Sapiens", which I just completed. I passed over the chance to read it many times before, but when I finally did I found it had a rare quality, in that no significant flaws in what was described lept out at me.

The combination of the "Imaginary Order" that Yuval describes, with the dynamics of Fragility vs Antifragility that Nassim describes seems all too fitting for the present moment. The imaginary order of the AI industry's highest valuation startup went up in smoke at a record pace, just short of 1 year after their release of ChatGPT.

Anthropic is essentially the same, as are all of the other cookie-cutter "let's just copy Google" trashbot companies. They build fragile systems, and are themselves fragile as companies, often compensating for that with cult-like imaginary order, whether it stems from the butchered ghost of "Effective Altruism", or the new brand of Death Kool-Aide that is "Effective Accelerationism". OpenAI's convoluted corporate structure was wholly unable to prevent this.

Out of all of the pundits making grand claims and predictions about the AI industry and OpenAI in particular, who bet all of their chips on "OpenAI will implode within less than 1 year of ChatGPT"? How many half-wits in Venture Capital can you name who claimed that OpenAI would go on to become one of the world's most valuable companies 10 years from now?

Part of what has made the past week so amazing is that it has been so spectacularly stupid. It is like the board of a company got so high on their own supply as to ask GPT-5 how to run that company, with the system prompt of "You are George R.R. Martin".

They stand as a testament to fragility, which any cybersecurity researcher should find poetic justice in. The king of the North and his shambling horde of P-zombies are still on the move, but his ending will come in due time.

Again, this begs the question "Will those funding AI Startups actually learn anything?". They've had their wake-up call, but they may hit snooze and continue on with more of the same.

Will you be P-zombie or Sapien, Fragile or Antifragile?