$100 billion sounds a lot, but in comparison it’s miniscule. For that amount of money you can either run an international space station for 24 years or the US military for 1.3 months, or the russian military for one year. If we’d just start killing eachother a little less, we could easily save that much money.

I would assume inflation made the number bigger today even if costs have dropped.

Right now? Effectively infinity dollars. Or Euros. Or Dongs. Regardless of the currency used, really. The only operational launch vehicles that have the diameter and payload capabilities to put up modules of comparable size are the SLS and Vulcan Centaur. Everything else comes up short in size, payload, or both. But getting the modules up there is only step 1, they also need to be attached. For the ISS, the Shuttle’s Canadarm was used to actually assemble the station in its early stages. Nothing currently operational has that capability.

Of upcoming launch vehicles, New Glenn and Starship are both capable enough in both size and payload. But of all the options, only Starship could be equipped with something like Canadarm. Starship is an orbiter with a payload bay. Everything else’s second stage is just engines, tanks and a payload adapter for simply deploying something into orbit.

You might ask “Why not design the modules to dock and assemble the station autonomously?” It’s a valid question, but it comes down to two things: complexity and mass. Every module would need its own full suite of guidance and control systems. And they’d be used only when a module is maneuvering to meet with the rest of the station. That’s a lot of mass wasted on something that only gets used once, mass that could’ve been put into making the module itself bigger. There’s also the issue of any leftover propellants and their residues on the module exteriors. It’s practically inevitable that some residue from the thrusters will end up on the connecting surfaces of the modules. A lot of the common orbital maneuvering propellants are toxic as fyuck and you don’t want any of it near unprotected people. And any residue could interfere with getting a firm seal between modules.

Another thought is to put something like Canadarm on the first module, but that also wouldn’t really work. Soon as it tries to move a second module into position, the two nearly equivalent mass objects would start tumbling around the common center of gravity, putting the entire proto-station into an uncontrolled spin on several axes. Stopping it would require, once again, thrusters.