He goes through his reasoning for why they took their $500M corporate treasury and put it all in bitcoin:
(1) Inflation is a vector, not a scalar. He buckets inflation into four groups: deflating goods (eg. computer chips, most mass produced goods, most digital goods), stable goods (eg. branded paper towels, branded consumer goods), elite luxury goods (eg. a Harvard degree, certain medical procedures, Picasso paintings), and financial assets (eg. treasury bonds). He claims that CPI is a headfake because it only follow the first two buckets. If you look at prices on elite luxury goods and financial assets, inflation is around 10% a year. He therefore concludes that holding $500M in corporate treasury creates effectively a $50M/yr negative cashflow, and they need to put that much into the treasury just to tread water.
(2) In 2010 a 10 year Treasury bond yielded around 5% interest, today it is 0.5%, a 10x reduction. He characterizes this as a 10x increase in the price of getting $100k/yr in zero risk yield: getting that much yield in 2010 would have cost $2M, but today it would cost $20M. That works out to around 10-20% inflation per year on that asset.
(3) He discusses why you need a corporate treasury in the first place, instead of giving all the money back to investors. He says it's like a rainy day fund, so the company can give assurance to its customers that they won't disappear in a recession.
(4) He goes through some other alternatives they considered: buying stocks (too illiquid for small cap, why is a public traded company buying other public traded companies?, inflated prices), buying treasury bonds (yields too low), buying gold (worse than bitcoin in every way)