Tom Chivers observes that the median expectation is that AI will boost the economy enormously, making shorting the market a terrible idea. However, he notes a non-trivial tail risk: that AI could kill everyone, in which case shorting would be pointless. This argument, Chivers suggests, may be almost deliberately missing the point.

Tyler Cowen—the George Mason University economist and Marginal Revolution blogger known for applying market-logic rigor to widely debated claims—responds that it is straightforward enough to buy seriously out-of-the-money puts while remaining long with the rest of your portfolio. Yet few, if any, of those who publicly worry about AI risk are actually doing so. Cowen points out that many—perhaps most or nearly all—of the catastrophic AI scenarios involve intermediate stages of significant disruption and panic, during which holders of such puts could cash in well before any worst-case outcome unfolds. No leverage is required; one could spend $5,000 or $10,000 a year on put options and, if wrong, treat the cost as a lapsed insurance policy. For those unfamiliar with finance, Cowen adds, an AGI could guide you through the process.

Some commentators argue that markets are poor at pricing long-term idiosyncratic risk—a concern rooted in the broader academic debate over whether equities systematically underprice distant, low-probability threats such as climate catastrophe or geopolitical conflict. Cowen counters that this is all the more reason to purchase those puts.

Cowen argues that when people attempt to rebut the idea that their stated views should logically lead them to short the market in some fashion, their reasoning tends to contain numerous errors in financial theory. He cautions against accepting claims about long positions as evidence of conviction, since long positions can still generate normal, risk-adjusted rates of return even if the underlying thesis is entirely wrong. It is shorts—whether explicit or implicit—that reveal an investor's genuine convictions.

Cowen references Victor Niederhoffer, the quantitative trader and author who held that an investor should never go net short on an asset. While Cowen is reluctant to use the word "never," he acknowledges his own view is not far different.

Cowen then poses a direct challenge: is it really so difficult to say, and act on, "Thank you, Tyler, I just went out and bought those puts"? Apparently it is. And perhaps, he suggests, that reluctance stems from an intuitive realization that while AI may impose significant costs, it ultimately will not be as catastrophic as some claim—a sentiment he describes as "words to live by."