First, take the "thorium is superior in almost every way" stuff with a grain of salt. There are pros and cons to any fuel choice from a technical perspective, and some of the "pros" offered up with regard to thorium are not unique to thorium, but just any Gen IV reactor design are compared to current gen reactors.

But the answer to, "why uranium?" is because of what we call path dependency: you make one decision (for one reason or another) early on, and it limits the ease of changing your decision later.

The initial nuclear infrastructure of the United States was uranium-based because uranium was necessary for the production of nuclear weapons during World War II. During the Manhattan Project, the US built up tremendous facilities for enriching uranium (for weapons) and built uranium-fueled reactors to produce plutonium (for weapons). After WWII it built more of essentially the same facilities in order to make more enriched uranium and plutonium (for more weapons). (It is an interesting historical aside that the US did investigate thorium for weapons purposes during the Manhattan Project, but concluded it would be difficult to use compared to plutonium, which itself proved more difficult to use than they expected, but once they had mastered uranium/plutonium for weapons, their interest in thorium waned.)

The Soviet Union essentially cloned the US approach to this, very deliberately, because it was known to have worked (aside from espionage, the US publicized quite broadly the basic approaches it took and their success). So by the early 1950s you have two superpowers heavily investing in uranium fuel cycles for their weapons program.

These two superpowers also were the first nations to explore reactors for power purposes. The US approach to reactors was spurred on by the needs of naval propulsion — nuclear reactors for submarines — which had very specific design specifications (they needed to be very compact). The reactor design they went with was one that used highly-enriched uranium fuel, because you can pack just a lot of power in a small reactor with that. And the US also had very easy access to highly-enriched uranium at this point, because of its weapons program.

That same design that the US used for its naval power reactor became the first template for civilian reactors — the Pressurized Water Reactor. And in the US, all uranium enrichment facilities were monopolized by the US government, who could set whatever price they wanted for the fuel that power companies might use.

All of which is to say: you can see that a nation like the US (and the USSR was similar in this respect) that has already invested trillions in a uranium-fuel cycle infrastructure for weapons purposes, and created reactors based on that infrastructure, is going to be a nation where any further nuclear reactor work is likely going to be on a uranium fuel cycle. Starting up the infrastructure for a new fuel cycle is tremendously expensive (aside from mining the raw materials, you need all of the facilities for milling the ore, plus facilities for doing whatever chemical and metallurgical transformations you need to make it into fuel, plus designing and proof-testing the reactor themselves, plus all of the regulatory infrastructure, etc.). Once you already have a massive uranium fuel cycle, then it's easy to stick with that.

Now, you might wonder, what about other countries? The answer is that most of them followed the same pattern, both because it was "tried and true" and because most of them developed their civilian programs with the assistance of the superpowers (and the US in particular offered fuels to other nations at a cheap price, because it didn't want them to develop their own fuel cycle infrastructure, because they thought that would increase proliferation potential). So the initial local path dependencies become global ones through the actions of the superpowers.

Thorium wasn't totally abandoned by the superpowers. The US and USSR both looked at it as a possible fuel cycle and for weapons purposes (both tested U-233-fueled bombs, which are made from the thorium cycle). US reactor researchers developed the thorium-based breeder reactor designs that are being so touted today (and even pursued by a few nations whose have ready access to thorium ores), and their possibilities for creating a fuel cycle that was almost totally regenerative was pushed strongly even in the early 1950s, before commercial reactors really existed.

But aside from the infrastructure needs, the positive aspects of thorium weren't really seen as all that important at the time. Breeder reactors in general promise nearly-unlimited fuel... but they were already in a situation of having enough fuel to meet their needs for decades. Breeder reactors promise less waste... but in the 1950s in particular, they weren't that concerned about waste issues. Even today, the fuel availability issue is not all that motivating, because it would require a huge expansion of the nuclear industry to run into fuel scarcity problems. But the waste issue is taken more seriously today, although high-level reactor waste is still not that large of a volume of waste (and most pro-nuclear people don't worry too much about it, which makes me suspect that their emphasis on it is just meant to allay the concerns of others, concerns they don't take that seriously in reality).

Anyway, as I said before, there are technical pros and cons with any fuel choice, and building up any infrastructure takes a lot of investment, so it is not that surprising that the one that has been focused on is the one that the nuclear nations had already built up beforehand, to meet their military needs. If the path of history had been a different one — e.g., power reactors before naval reactors and bombs — perhaps a different choice might have been made, but if your path is bombs to naval reactors to power reactors, then the uranium cycle is probably the one you'd end up with almost by default.