Generation 1
The ChatGPT Era · 2022–25
Free / Constrained
Gen 1a
Whatever your organisation provides by default. Shorter context, weaker models, no memory. Limited by design.
Free ChatGPT, free Claude, basic Copilot, bundled tools
My strong recommendation: don’t waste time here
Pro / Competent
Gen 1b
Good general-purpose applications. Dramatically more capable than the free tier — longer context, stronger models, useful features.
Paid versions of Claude, Gemini, or ChatGPT
My strong recommendation: Claude
Generation 2
The Claude Code Era · 2026 +
Agentic / Frontier
Gen 2
General-purpose agents that can interact with your computer. They don’t just answer questions — they read files, write files, pursue objectives, and complete entire workflows independently.
Claude Code, ChatGPT Codex, Google Gemini CLI
My strong recommendation: Claude Code
Gen 1 → GPTs
Packaged prompts
Productised as GPTs — packaged prompts for single tasks. You work out how to research a company, save the prompt, reuse it for the next one. One task, one prompt, one output.
Gen 2 → Skills
Packaged agents
Productised as skills — packaged agents for whole workflows. I built a process that researches every nightclub in a market, scores them on multiple dimensions, and produces a live dashboard. Then reran the whole thing in one click for completely different industries:
1
For the many
Pro / Competent · Gen 1 tools
Free Pro
Move less-engaged people from constrained, default tools to competent use of good general-purpose applications. This in itself is a big return on investment.
2
For the best
Agentic / Frontier · Gen 2 tools
Pro Agentic
Accelerate your best people with agentic tools. They’ll build workflows that deliver outsized impact — and those workflows become reusable for everyone else.