Wind isn't the lowest cost for baseload - if it isn't subsidized. It's not even competitive with natural gas, and is markedly less reliable.
It's subsidization which makes it cost competitive. And even then, it's only competitive per kwh, it's not competitive for primary base load (which is 100% what a datacenter needs) because wind is cyclical and periodic. It isn't always windy in WY, and it's often too windy for wind power (35mph+ winds). I'm really tired of the tripe propaganda about wind/solar. I LIKE wind and solar, conceptually (and for the ability to run it off-grid), but let's be real.
The solution here, long term, is likely SMR generation at-scale. You've got many datacenters, and the capacity scales rapidly. A single large reactor doesn't make fiscal sense, but a national program to produce industrialized SMRs at scale which could be deployed as needed over a period of years would, enabling cheap power generation. When you build at scale, you're able to drastically drive costs down.
Fission is also now on the horizon.
But in the interim, green coal and NG are likely to be the thing.
(I'm not looking forward to the massive impact that this is going to have on the regional NG market; NG has already gotten significantly more expensive, and so many people use it for heating.)