AI has been adopted faster than any other technology in history. In ~4 years, AI went from an idea to a product that more than 1.2 billion people have already used. In the US alone, 1 in 4 adults now interact with AI every day, with worldwide AI spending expected to hit ~$2.6 t
Why now
Thesis: AI adoption narrative underpins long-run compute spending
Catalyst to watch: Next earnings capex guidance from hyperscalers
Main risk: Spending forecasts may prove inflated
Why it matters
Adoption and spending figures frame AI as a durable multi-year capex theme, supporting compute and infrastructure demand.
Details
AI has been adopted faster than any other technology in history. In ~4 years, AI went from an idea to a product that more than 1.2 billion people have already used. In the US alone, 1 in 4 adults now interact with AI every day, with worldwide AI spending expected to hit ~$2.6 trillion this year. But almost all of that AI lives on a screen, as chatbots answer, draft, advise, then hand information directly back to its user. This still requires a substantial amount of human interaction, effectively bottlenecking efficiency. For example, despite AI agents, the average American still spends 41 minutes on hold every month, which adds up to 2.8 billion hours per year nationwide. The next phase of AI will be linking it to the real-world in a far more efficient way.
Sources
- x · 2026-09-28 18:32 UTC