A $3 trillion gamble and global diffusion: The bipolar narrative of AI in 2026

After reviewing the 2026 trend outlook reports from five top institutions—a16z, Goldman Sachs, JPMorgan Chase, Morgan Stanley, and BlackRock—I extracted two key value points:

1) What bubble? Will the AI industry enter a period of accelerated investment?

Morgan Stanley has given a staggering figure: capital expenditure on AI infrastructure is projected to reach $3 trillion, with less than 20% currently deployed.

What does this mean? Amazon, Google, Meta, Microsoft, Oracle, and other hyperscale cloud providers are now spending a fortune building data centers, buying GPUs, and laying power infrastructure, but this is just the beginning.

However, JPMorgan Chase offered a sober assessment of the actual benefits of such large-scale AI adoption, believing that in the short term it can only boost the profits of some companies and help giants optimize their profitability. It will take many years to truly achieve the significant benefits of a qualitative leap in AI productivity.

In fact, it only made one point: 2026 will still be a year of crazy spending on AI, but it is still just the investment period and far from the time of harvest.

2) US stock market concentration dividends and non-US market spillovers: which side are you on?

BlackRock has proposed a concept called "Micro is Macro," which argues that AI investments by a few companies already have a macro-level impact.

Data shows that in 2025 (YTD), the equal-weighted S&P 500 in the US stock market only rose by 3%, but the market capitalization-weighted version of leading technology companies rose by 11%. This 8% difference may be due to the benefits of AI concentration.

Morgan Stanley is the most aggressive in this regard, setting a target of 7,800 points for the S&P 500, which represents a 14% increase from the current level, based on the assumption that the profitability of the seven tech giants will continue to strengthen.

However, JPMorgan Chase believes that as the dollar weakens, the benefits of AI will spill over into the global supply chain, thus giving emerging markets an annualized expected return of 10.9%, higher than the 6.7% for US large-cap stocks. Goldman Sachs also sides with the spillover effect, giving emerging markets the same 10.9% expectation, believing that Europe has the potential for 7.1% and Japan for 8.2%.

Simply put, these are two completely different bets: BlackRock and Morgan Stanley are betting that the AI dividend will continue to be monopolized by US tech giants, while JPMorgan Chase and Goldman Sachs are betting that AI is a global infrastructure upgrade, and the dividends will spread to global non-US markets.