Genius is one percent inspiration and ninety-nine percent perspiration."
— Thomas Edison

this explains AGI, because AI models greatest stregenth is perseverance and 24/7 work. even Edison slept a few hours a night when he was at his most productive.
the historical capex booms that failed spectacularly (railroads, Dotcom/telecom/shale) were fueled by debt and ipo dollars way before they had real revenue/gross margin/cash flow. of course, long term they were great investments. what is different this time is that the frontier model companies and hyperscalers are funding the ai Capex boom from cash flow with some debt and gushing revenue. the bubble flush happens when there is a revenue lag between buildout and revenue.
I remember the QCOM 1999 rise well. One thing to note as a difference is how much of many of the exponential rises were driven by almost pure multiple expansion (QCOM and MSTR come to mind) versus earnings plus multiple expansion. SNDK definitely in the latter camp. The exponential rise in earnings coupled with a re-rating from a commodity market multiple based on short-term purchase agreements to a more stable long term grower based on long term purchase agreements argues for a differential analysis.
@JesusFerna7026 yeah the ultimate hedge is retirement age, if markets are bad keep working. given that 2-3 years usually sees you through to the other side it’s not too much to ask. real estate is another good hedge. if holding in a good market, steady cash and no real terminal value risk.