Archive · 6 Sep 2026
AI Investment Bets Face Questions Over Slower Economic Adoption
Technology companies are spending a huge amount of money to build AI infrastructure.
They expect people across the economy to use AI very often.
Software engineers already use AI coding tools regularly, so they provide the strongest evidence for this idea.
Other workers may use AI less because people must still check its work and take responsibility for mistakes.
AI is entering fields such as law, medicine, finance and marketing.
However, occasional use by these workers may not require as much computing power as constant use by programmers.
The technology could therefore become widespread without producing enough demand to justify all the infrastructure being built.
This does not necessarily mean AI will fail.
It could instead mean that investors have built for faster growth than the economy can deliver, creating a possible infrastructure bubble.
Technology companies are committing hundreds of billions of dollars to AI data centres, chips and electricity.
Software developers are currently the clearest example of intensive AI use, especially through coding assistants.
Lawyers, doctors and executives may adopt AI more cautiously because humans remain accountable for serious mistakes.
AI could spread across many industries without generating the same computing demand as software engineering.
The central risk is a mismatch between rapidly built infrastructure and more gradual economic adoption, potentially creating an infrastructure bubble.
- Who
- Technology companies, software developers, investors and workers in fields adopting artificial intelligence.
- What
- Technology companies are making enormous AI infrastructure investments based on expectations of widespread, intensive use, while the strength of that demand remains uncertain.
- When
- The article describes the situation as AI is currently spreading beyond software engineering.
- Where
- Across the wider economy, including law, medicine, finance, marketing and administration.
- Why
- Companies are investing because they expect AI to become an indispensable, heavily used utility across the economy.
AI Growth Case
AI Infrastructure Concern
Future demand
AI Growth Case
AI is already becoming a routine tool for software developers and is spreading into law, medicine, finance, marketing and administration.
AI Infrastructure Concern
Adoption across industries may be widespread but relatively light, generating less computing demand than technology companies expect.
Economic productivity
AI Growth Case
AI coding assistants can help one worker complete tasks that previously required considerably more time, suggesting substantial productivity potential.
AI Infrastructure Concern
The productivity gains may be harder to realise in occupations where people must carefully supervise AI and remain accountable for its errors.
Investment outlook
AI Growth Case
If AI becomes an indispensable and heavily consumed utility, the large investments in data centres, chips and electricity could support an industrial transformation.
AI Infrastructure Concern
Infrastructure may be built faster than demand develops, creating a mismatch that could make current AI investment resemble an infrastructure bubble.
- Investment scale
- Technology companies are spending hundreds of billions of dollars on data centres, specialised chips and electricity.
- Strongest current use case
- Software developers increasingly use AI to generate, debug and rewrite code.
- Other adopting sectors
- AI is entering law, medicine, finance, marketing and administration.
- Accountability concern
- Lawyers, doctors and executives cannot simply transfer responsibility for consequential mistakes to AI systems.
- Demand distinction
- Widespread AI adoption may not produce the same computing demand as intensive use by programmers.
- Main risk
- Infrastructure could be built for explosive demand while the economy adopts AI more gradually.
- Possible outcome
- The AI boom could become either an industrial revolution or an infrastructure bubble.