Elon Musk Responds to Atreides CIO's "Compute Inequality" Warning as Firm's AI Spending Doubles Monthly

Over the weekend, SpaceX CEO Elon Musk responded to a prediction from Atreides Management Chief Investment Officer Gavin Baker. Baker predicted that rapid growth in AI usage will make compute spending increasingly critical.

Atreides AI Spending to Grow 100-Fold

Baker posted on X that Atreides' internal AI spending in August 2026 will be roughly 100 times higher than in March, while continuing to grow at approximately a monthly doubling pace.

He stated that this growth stems from deep integration of AI into investment research and other knowledge-intensive work.

Baker expects the firm to find a balance between "compute-efficient humans," cheaper open-source AI models, and more expensive frontier models.

Baker Warns of "Compute Inequality"

"AI means scale matters more for investing," Baker wrote on X, adding that there may eventually be a minimum token spending threshold required to remain competitive in many knowledge-intensive industries.

He described this potential gap as "compute inequality," suggesting that organizations with resources to consume more AI could gain advantages over those that cannot.

Baker also noted that his spending could rise further when Grok Bot shifts to a consumption-based pricing model, which he expects could create a "step function" increase in his AI costs.

We expect to start hearing about "Pareto optimal" balances of compute-efficient humans, cheaper open-source tokens, and frontier tokens.

Our internal AI spend at @Atreidesmgmt will be ~100x higher in Aug 2026 vs Mar 2026. Still roughly doubling monthly...

— Gavin Baker August 23, 2026

Elon Musk Responds

Musk, who owns AI company xAI and its Grok chatbot, shared Baker's post and said "Interesting."

Interesting

— Elon Musk August 23, 2026

Larry Fink Warns AI Could Create a "K-Shaped Economy"

In May, BlackRock CEO Larry Fink warned that the massive costs of adopting AI could sharply differentiate market winners and losers. "As the AI economy enters every industry, we will see a K-shaped economy," Fink said.

He argued that the massive capital investments required to compete in AI could make this divergence increasingly permanent, as companies race to secure the infrastructure needed to remain competitive.

Fink highlighted growing shortages of electricity, semiconductors, and data processing capacity, predicting that "a new asset class will be buying compute futures."