Stupider and Stupider
If what’s marketed as AI has value, it will be in opening humanity to the understanding technology is politics. This remains to be seen, but there’s a number of recent optimistic sightings. Here’s a nice short interview on the Tech Report with a guy named Paris Marx who has put out a book, Hyperscale: The Ambition and Excess of Big Tech's Data Empires. It probes some of the PR idiocy that’s helped float the AI bubble of the last few years.
Speaking of PR idiocy, (Mis)Anthropic released their initial paper work for an IPO. Why not? After all the SEC let Spacex and General Fusion go in the summer. There’s little to say about what Anthropic released. It’s pretty much all fantastical. Numbers, just like words, can be used to shape all sorts of fantasies.
Speaking of fantasies, General Fusion, the energy hallucination of Jeff Bezos, is now 60% below its IPO price high of a couple months ago. Why it's even that high I can't tell you. I’m told Bezos says this is ok because in 20 or so years they’ll be making all sorts of money, once they get the plasma thing worked out.
General Fusion is just the greatest example of the nonsensical, mad nuclear rush of the last several years. How can there be artificial intelligence when natural intelligence seems so sparse? The Valley’s coders convinced themselves splitting atoms was not much harder than an html script. Typical is a company called Nuscale, “putting a nuke in every backyard,” which is now just 10% of their year ago high. The whole nuke sector tells the same story, keep in mind, no electricity, no trillions of dollars of centralized compute infrastructure. That would be an unmitigated good thing.
The question of some sort of financial reckoning is a difficult one. Finance today is not the same as it was forty or even twenty years ago. There’s a number of reasons, but two would be concentration and the automation of trading. Few talk about either of these, but how this impacts markets is essential to any understanding of how markets supposedly function or better, dysfunction today.
A third factor is a lack of regulation, Spacex, General Fusion, and now it seems Anthropic IPOs being prime, but by no means exhaustive examples. If the entire system provides any value or as we’re incessantly assured a means of measuring value, the best comparison might be the literal Wild West of 19th century US finance, except now with the Fed providing a backstop.
Hard to see how a shakeout in AI finance would not severely impact the entirety of finance. What’s important to contemplate is when/if things really start to shake, the Fed will be handing out trillions. This is politics and there is no organized political opposition to concentrated finance, after all, why should the 90% have any say in the affairs of the 10%? The Fed's actions, just as they did in 2020, 2008, well since its inception, will lead to an even greater concentration of wealth, along with this time a lot of compute to be had on the cheap, leading to greater concentration of compute power, think here MS, Amazon, and Google – not a good thing.