David Booth: AI may change the world. It just won’t change how you invest

The article argues AI is more likely to improve productivity across companies (e.g., data centers/chips with $1.2T+ expected capital spending in 2027) than to help investors beat stock and bond markets reliably. It emphasizes market efficiency—AI-derived information gains are shared by all participants—casting doubt on “AI stock-picking” or concentrated bets in AI giants. Overall, the message is a cautious stance that AI may enhance innovation, but investors should avoid overconcentration and expect uncertainty rather than market-beating returns.

Analysis

The immediate market implication is not about AI creating alpha; it is about AI compressing the value of everyone who sells judgment. If model access becomes cheap and widespread, pricing power migrates away from stock pickers and toward firms with proprietary data, distribution, and embedded workflows. That is structurally constructive for scaled platforms like GOOGL, but only if AI translates into measurable monetization rather than just higher compute intensity.

The 1-3 month risk is a capex-versus-revenue mismatch: AI spend can keep rising even if downstream monetization is delayed, which leaves high-multiple software and cloud names vulnerable to multiple compression if earnings revisions lag. Over 6-18 months, the more interesting beneficiaries may be non-tech operators that use AI to cut inventory error, labor hours, and marketing waste. That second-order effect would favor companies with obvious working-capital leverage, but only after the operating data proves it.

The contrarian point the market may be missing is that democratized AI can reduce informational asymmetry faster than it creates new stock pickers, which makes broad exposure more attractive than concentrated "AI winner" bets. The falsifier is clear: if the next two reporting cycles show durable revenue acceleration or margin expansion tied to AI, then the thesis that AI is merely table stakes is too conservative. Until then, this reads as a dispersion trade, not a single-name breakthrough story.

Key Decisions for Investors

• Hold GOOGL as core AI exposure, but do not add on narrative alone; reassess after the next 1-2 earnings prints for evidence that AI is converting into incremental revenue rather than just higher capex. Trim if cloud AI monetization or ad growth fails to inflect.
• Pair trade for 1-3 months: long QQQ / short IGV. If AI is commoditizing analysis and compressing moats, broad megacap exposure should hold up better than high-multiple software names; cover the short if software growth reaccelerates or multiples de-rate less than 10%.
• Treat LEVI as a watchlist beneficiary, not a current trade: only get constructive if the next 2 quarters show better inventory turns or gross margin stability, which would validate AI-driven planning benefits. Otherwise, no position.
• Avoid paying up for concentrated 'AI alpha' narratives in active-manager or research franchises until there is proof of persistent outperformance; if the next cycle shows no skill persistence, fade fee-bearing stock-picking products rather than the index.