A Rather Dubious Bloom: AI & the Market

Palantir, my dears. Such a name. One imagines a gothic novel, not a purveyor of data analytics. They’ve taken to calling themselves an “n of 1,” which, frankly, sounds less like innovation and more like a particularly tiresome bout of self-regard. Their stock performance this year suggests “n of negative 1” might be closer to the mark. A valuation at 128 times earnings? Really? One begins to suspect the emperor has not merely no clothes, but a rather glaring lack of common sense.

The question, of course, isn’t simply whether Palantir is overpriced – though it demonstrably is – but whether one might find a slightly less… optimistic investment in this current AI frenzy. Let us consider a couple of possibilities, shall we?

Nvidia: Not Entirely Without Merit

Nvidia, unlike our friends at Palantir, doesn’t quite induce a migraine upon sight of its valuation. At 24.5 times forward earnings, it’s almost… reasonable. Almost. The impending Rubin platform – due in 2026, which in market years is practically the next millennium – promises to reduce inference costs and boost training capabilities. One suspects, however, that much of this potential is already factored into the price. Still, one can’t entirely dismiss the prospect of continued growth.

Loading widget...

The talk of an AI bubble is, naturally, rampant. Nvidia’s Mr. Huang, with a characteristic air of confidence, insists it’s different this time. One hopes he’s right. But let’s be clear: demand for powerful chips will likely continue, but that doesn’t preclude a rather spectacular correction along the way. Nvidia, for the moment, appears best positioned to benefit, though one wouldn’t wish to stake one’s entire fortune on it.

Advanced Micro Devices: A Challenger, of Sorts

If one feels compelled to dabble in GPU stocks – a rather reckless impulse, in my opinion – one might consider Advanced Micro Devices. They’re the most credible challenger to Nvidia, which is a polite way of saying they’re the only ones even attempting to compete. At nearly 32 times forward earnings, they’re still not exactly a bargain, but they’re considerably less… audacious than Palantir.

Loading widget...

Their Instinct MI400 chips, they claim, will give Nvidia’s Rubin a run for its money. More memory, more bandwidth – all very impressive on paper. The hyperscalers, naturally, are anxious to diversify their suppliers. One suspects this is less about genuine competition and more about leverage. Once the MI400 hits the market, AMD’s stock may enjoy a temporary boost, but whether it can sustain momentum remains to be seen.

Software vs. Silicon: A Rather Predictable Debate

Normally, one would assume software will ultimately triumph over silicon. It’s more scalable, more adaptable, less subject to the whims of manufacturing. However, that assumes comparable valuations. Palantir has a perfectly adequate product, impressive growth, and a strong underlying business. Unfortunately, its stock is priced as if it has already solved world hunger. No company can sustainably live up to such expectations.

Nvidia and AMD, while hardly immune to market fluctuations, at least have the advantage of being grounded in reality. They don’t need to be perfect to deliver returns. And in this particular market, “good enough” is often more than sufficient. One should simply approach with a healthy dose of skepticism – and a very large glass of something bracing.

Read More

2026-02-16 12:53