chat

I have no term for this other than zombie economy. Everything's dead and rotting but keeps moving. It's so baffling how much it's been pushed when it's provably not good and lies all the time. I saw an article talking about like a report where some senior developers used LLM coding tools and self-rated their productivity as higher when in actuality it was 20-25% worse. All of that time was a result of having to double check everything the programs did because they were wrong so often. Such a strange phenomenon. Like, if we all really believe it works it will surely do so eventually, right? Is this idealism?? Magical thinking??? But it's like the machine casts spells on us, instead. Purely because the output reads as kind of human sometimes.

Keep in mind that all these financiers and politicians are (for the most part) the same people who cashed in or bailed on the internet in 99. The smart ones know they're building another dotcom bubble but they don't want to lose the game of chicken at this point.

I mean the real ghouls are going to be making hay when the correction does come, the house always wins etc. etc. But I think the spectre of dotcom and then SF finance cashing in on web 2.0 makes up the superstructure here.

Crypto is big, but it never quite got the wide adoption its proponents were aiming for. Never mind that it has a market cap of $4T or whatever, that’s just crypto weirdos trading with one another and pretending that trading prices mean value. The amount of liquidity available to turn crypto into usable cash is minuscule.

The thing that crypto has going for it is that it was never designed to do anything useful. Its entire value is in being a store of value that is valuable because people treat it as a store of value. It has nothing else to prove to anyone for it to maintain itself. Whether it continues to maintain its value is questionable, but it’s hit a point of modest stability because it did the job of getting a bunch of retail investors to buy in and hold forever, which was the goal, and that situation can last as long as those retail investors continue to diamond hand their assets en masse.

AI isn’t in the same position. It’s not enough that lots of people are telling everyone that AI is good, or that big dollars are going into it. To be successful, AI has to eventually turn into surplus dollars. In order for it to do that, people have to pay for it, and on the order of hundreds of billions of dollars a year. I’m not saying that won’t happen, because who knows? The market can stay irrational for a very long time. But the market does eventually demand returns, and those returns depend on a lot of factors that are currently not manifesting.

Basically, crypto is stupid, but it makes sense in that it serves the intended purpose. It’s still a bubble, but it’s a bubble that can stay inflated for a long time because of the nature of the asset. AI is a bubble that needs a lot more cash to sustain. If it is to be sustained, it’ll likely be governments writing increasingly large checks for useless technology in perpetuity, which is a real thing that could happen. It’s very, very stupid, but the alternative is that the line go down, so hard to say which is more likely to happen.

wheresyoured.at is Ed Zitron's substack. His most recent free post, "A Hater's Guide to the AI Bubble" is a good place to start.

Like I said, Ed Zitron is a good source. His newsletter Where’s Your Ed At? is pretty thorough, and his podcast Better Offline is more of the same, but in pod format. He writes angry (which may or may not appeal) but he brings receipts and does a lot of breaking down specific arguments around the tech industry, and more recently AI specifically.

On the specific point of the AI industry needing to be bigger than smartphones and cloud combined (I think it might have been smartphones and SaaS combined, but the point is that it’s ludicrous) it’s a pretty straightforward matter of the amount of capital invested. Hundreds of billions of dollars are going into AI. For those investments to pay off, the AI industry needs to be making hundreds of billions in revenues. The smartphone industry is ~$800B in revenues last I checked. The AI industry is ~$35B with massive losses, and those revenue numbers are very suspect because of all the inside baseball nonsense between all the big tech companies.

They’re talking about investment in AI tooling a trillion dollars before the end of the decade. That simply requires that the AI industry be worth quite a bit more than that by the time the money gets spent. The specific numbers are less relevant than the fact that the broad numbers aren’t close to making sense.