Buyer's market 🤝, AI productivity fails 📉, the biggest moat 🌊

AI productivity gains

Most AI users are only 10-20% more productive despite 'game-changing' claims; real transformation requires organizational refactoring, not just tool use.

Attacking competitors publicly

Public competitor attacks never work because naming a competitor only gives them more exposure and most people don't care about inter-company disputes.

OpenAI and Anthropic consulting joint ventures

The author bets against new OpenAI and Anthropic consulting joint ventures on the grounds that most enterprise data isn't ready for the work being sold.

AI chatbots

Chatbots should be deprecated because they essentially offload interface design work to users.

Enterprise AI agent adoption

Of 50 Midwest enterprise CIOs, no one had agents at scale and only 5 of 25 had agents in production at all — the bottleneck is undocumented workflows, not model quality.

Use AI Slop as Raw Material, Not Final Output

The 'slop cannon' approach treats junky AI output as raw thinking material to edit hard rather than polished deliverable — move fast in produce mode, slow down in selection mode.

Reframes AI slop from insult to strategy — using low-quality AI output as a speed-thinking scaffold rather than avoiding it entirely.

Momentum as a Moat Against AI Disruption

Rather than building traditional moats like economies of scale, companies should prioritize rapid execution and cohesive shipping speed to outpace larger incumbents.

Anti-conventional in that it argues speed and momentum beat traditional structural moats in the AI era, even against giants.