Two Kinds of AI Users Are Emerging - Martin Alderson

2026-05-14 · source · source ↗ · raw capture · wiki page

Main Thesis

Dramatic divide between AI power users (Claude Code, MCPs) and casual users (basic chatbots). This gap increasingly determines organizational productivity.

Power User Characteristics

Often surprisingly non-technical professionals
Adopt cutting-edge tools across diverse roles
Leverage programming ecosystems for complex tasks
Experience transformative productivity gains

The Microsoft Copilot Problem

"M365 Copilot has enormous enterprise market share yet feels like a poorly cloned ChatGPT interface." Microsoft internally adopts Claude Code despite owning OpenAI.

Enterprise Barriers

Locked-down environments prevent script execution
Legacy systems lack internal APIs for agent integration
Siloed engineering departments can't build safe infrastructure
Security concerns create genuine bottlenecks

The Productivity Gap

Smaller companies without corporate constraints experience exponential advantages. Example: converting a complex 30-sheet Excel financial model to Python "almost one-shot" with Claude Code.

Key Conclusions

Organic adoption works better than top-down AI strategies
API-first architecture becomes crucial competitive advantage
Legacy SaaS becomes problematic -- lock-in or bottleneck
Future: agents connecting to systems via APIs, democratizing knowledge work