Bradley Tusk 解释他为何认为“叙事经济学（narrative economics）”可能正在推动当今的 AI 基础设施大爆发： “我至少有一种怀疑：目前这一轮巨额的 AI 基础设施投资——2025 年规模 1.5 万亿美元，预计 2026 年将达到 2 万亿美元——其实更多是为了 AI 大公司的短期股价和估值服务，而被包装成‘长期思考’。 投资者喜欢两样东西： a.) 公司自身的基本面； b.) 但往往更喜欢的，是叙事，对吧？ 如果不是因为叙事，以及投资者相信 Elon 能搞定那些连他现在都无法预测的事情，那么 Tesla 今天的估值连现在的一小部分都不可能达到。我的意思是，从任何客观指标来看，丰田（Toyota）都应该比特斯拉价值高得多，但上次我查的时候，它的市值还不到特斯拉的四分之一。 所以我确实有点担心，目前发生的大量投资，其真实目的其实是推动叙事，进而提高私有公司的估值，或是提升上市公司的股价， 而不是基于未来 10 年、20 年经济真正需要什么。” Bradley Tusk explains why he thinks 'narrative economics' might be driving today's AI infrastructure boom “I at least have a suspicion that all of this massive AI infrastructure investment - so it's $1.5 trillion in 2025, projected to be $2,000,000,000,000 in 2026 - is really about the short term share price & valuation of the big AI companies disguised as long term thinking. Investors like two things: a.) would be the fundamentals of a company itself, but b.) oftentimes even more so, is the narrative, right? Tesla should not be worth a fraction of what it's worth today, if not for the narrative and the belief by investors that Elon will figure things out that no one can predict, including Elon, right now, and therefore the value is worth it. I mean, Toyota should be an infinitely more valuable company than Tesla, but last I checked it was like less than a quarter. So I do worry a little bit that a lot of the investment that's happening is really about driving the narrative and therefore the valuation if it's private or the share price if it's public today, as opposed to what the economy will actually need in a decade or two.”