@_clarktang The problem with his kind of thinking and the MIT study is the non separation of production vs experimental use cases It is okay for the average or even great majority of AI experiments to not yield positive ROI. What is an ever expansion of wildly successful deployed use cases.
@altcap Good discussion. Debate about frontier models vs opensource models is basically the same as the one about nvidia gpus vs asics. Both focus narrowly on market share split and ignore how fast the pie is growing.
iOS once had 100% of the smartphone market. Don’t ever forget that.
@Tablesalt13 G7 has always meant “club of friends of the United States”. when you are not a friend and has no real economic and geopolitical might, you will be out.
@pbeisel it will be good for tesla and spaceX to eventually merge—but the timing is not now because Tesla is just about to be re-rated higher due to robotaxi and Optimus. Probably as much as 50% higher.
@GavinSBaker Expansion into the U.S. also gives the U.S. a buffer to act should China chooses to blockade the strait of Taiwan in the manner that Iran has blockaded Hormuz—the reverse blockade by the U.S. has only been possible because the U.S. has excess oil and gas supply.