Agree with most here that it's a puzzling decision based on valuation alone.
But there are other factors we're not privy to, of course, and over the long term I think owning a chunk Alphabet is better for shareholders than having $10B in cash today.
Just speculation and rumination, but it'll be interesting to see if this is Berkshire "picking their horse" in the AI capex boom. Not to say they are making a prediction about an ultimate winner or loser, but I remember at last year's AGM, Greg was adamant they would not let data centers hook up to BHE without the implicit understanding the data center(s) wouldn't receive preferential treatment, and that rates would not be increased for smaller customers due to capacity constraints. My feeling is Greg (and probably Warren too) feel more comfortable about how Alphabet's management is approaching these issues than the other players, but again, that is just speculation.