Source item

I have a simplified conjecture about AI progress. Your mileage may vary but here it is. f(i) = p × c × a where i=intelligence, p=power, c=compute, a=algorithms, multiplied together. This means neglecting any variable creates a ceiling and any breakthrough compounds across all three. One caveat is that today's advances are more accurately shaped by the entire history of prior investment, not just a discrete investment or breakthrough in time. In other words, the variables have memory over time so that must be factored in. In any event, the chart below is interesting because it shows the major tech companies operationalizing this function… If f(i) accelerates quickly, there is a point in f(i) where the gap between the haves and have-nots is so stark that negotiating leverage will flip instantly. This is the bet that Amazon, Microsoft, Google and Meta are making. This is why they are buying as much of p, c and a as possible to make sure they own the leverage that comes from f(i). Meanwhile Apple is cutting capex. Apple's implicit assumption is: “AI capability is an input we can buy, not a moat we need to build. We own the distribution and last mile so let’s wait and license the intelligence when a winner is more clear. It’s a feature that needs distribution not a platform that reallocates winners and losers.” It's a rational bet today but it's also the kind of bet that can look obvious until the day it isn't.