Investors once cheered AI IPOs as inevitable, but recent analysis shows many are statistical long shots. A review of recent listings reveals that, when subjected to a standard probability test, most tickers look more like speculative bets than solid growth stories in the market today.

Morgan Stanley’s Michael Mauboussin and Dan Callahan dissect the hype in their piece, “Bayes and Base Rates.” They argue that the scale of capital flowing into AI—surpassing historic railroad and internet booms—depends on sales growth that never has precedent within the next decade, investors expect returns that could outpace traditional sectors, but the uncertainty remains high.

The argument echoes the dot‑com era, where a handful of stocks survived, but most vanished. Comparing the scenario to a Central Park temperature forecast, the author notes the event has not occurred in 60 years, making the odds minuscule even with climate change in play, and the risk profile is significant.

For portfolio managers, the takeaway is clear: treat AI names with the same caution applied to high‑yield speculative ventures. Weighting these equities against proven fundamentals will shield against the steep losses that accompany over‑valuation, especially during volatile cycles when market sentiment shifts abruptly and investors.