In "Situational Awareness: The Decade Ahead", Leopold Aschenbrenner estimates that the largest training clusters will cost hundreds of billions of dollars around 2028.
Clarifications:
$100B worth of computers and associated data center infrastructure (e.g. building, cooling, networking; does not include the cost of the power plant)
Computers must be networked together and training one model. They do not need to synchronize weights at each gradient step.
Value of data centers will be estimated with reasonable depreciation. So, $100B of purchase price in TPUv1s would not count.
Nominal dollars.
So, if the nominal value of the datacenter being used to train a model less depreciation ever crosses $100B, this market resolves YES.
This is one of a series of markets on claims made in Leopold Aschenbrenner's Situational Awareness report(s):
Other markets about Leopold's predictions:
1,000
1.00Effectively. Nominal dollars at the time of potential YES resolution. That is to say, if on any day between now and 2029, there is a training cluster worth $100B of that day's dollars, then the market resolves YES. So yes, this market is affected by inflation of dollars but also depreciation of hardware.