I believe them: if anybody knows mediocre worthless software, it’s Microsoft.

Honestly, I expect this AI bubble to implode with much more devastation than the dot-com bubble.

And it’s not even that AI is useless. Like the internet (during dot-com) it will definitely have good uses.

But (a) those uses will take many years to crystallize and mature and (b) the early capital-intensive movers have a big disadvantage and most of them don’t have a feasible path to recoup the money invested into them.

This is why the AI club is licking Trump’s boot. They will get the federal government to bail them out by buying overpriced AI products and services and taking over worthless investments “in the interest of national security”.

American taxpayers are going to foot this bill and they will not like it when they start seeing the effects.

The highest population growth rate in recorded human history was sometime between 1965 and 1970, when the population was growing at about 2.1% annually. It’s down around 1.1% now. 10% hasn’t been realistic since we went from 10 humans to 11 humans.

(Note: I go on research benders. This particular rabbit hole (pun intended) became really interesting from a terrifying sci-fi perspective; kind of like a reverse Children of Men. As I wrote it up, it started to feel like an apocalyptic xkcd What If.)

So, think about the sheer numbers–for a 10% growth rate, we’d need more than eight hundred million births per year. There are only 1.9 billion women of childbearing age on Earth, which means that a little under half of every woman between 15 and 49 would need to be pregnant every single year in order to make that target (a little more than that to account for women who aren’t able to be pregnant for whatever reason, a little less than that to account for multiple births, let’s just say those will wash out).

Let’s imagine that this happened this year. Right now, we have about 132 million births annually. That would mean that we’d need about seven times the number of maternity wards, fully staffed. Five years after this unprecedented baby boom, we’d need to begin increasing the number of classrooms worldwide by a factor of 7 as well, which would mean that seven times the number of college students would have to go into education this year. We’d need to massively ramp up food production, probably starting sometime in the late 90s. We’d need to massively scale up infrastructure and housing on a Chinese scale. In 2050, our annual birth rate would equal our 2025 population, meaning we’d be adding an additional everyone who’s here right now every single year.

Interestingly, this would cause a precipitous and sustained spike in our death rate, since 0.2% of women worldwide die during childbirth. This means 1.6 million women dying in childbirth every year between 2025 and 2040 (which almost exactly balances out the number of women coming to childbearing age during those years to replace them), or the first couple months of COVID happening perpetually, but for women in childbearing. Every woman would have, on average, seventeen children by age 49, and if you know 33 women under the age of 49, one of them will die in childbirth.

In 2040, we’d have to start ramping up production again as the ten percent growth rate would begin forcing (because that’s obviously the only way for it to work) half of the 400 million women born in 2025 to start having their own children. Fifteen years later, do it again. Fifteen years later, do it one more time.

But you can probably stop worrying about it by that point. Studies are incredibly mixed on the total carrying capacity of Earth, ranging from 2 billion to 1.024 trillion, but a 10% growth rate would get us to even the most gigantic estimates within 50 years.

At some point, the pendulum begins swinging the other way; whether due to famine, water shortages, pandemics, climate change, or any number of other factors, there will be a huge increase to the global death rate once again. Our population would stabilize at our global carrying capacity; and from that point on, Earth’s line can never go up again without the line going down first.

It’s so patently ludicrous. Just total nonsense.

“I wasted a shit-ton of money and boiled the earth for epic AI profit and all I got was this lousy podcast”