Is giving employees ChatGPT, Claude and Copilot an AI strategy?
Companies spent tens of billions on generative AI and 95% saw no return. A licence is a purchase. A strategy decides what work to do and where a wrong answer costs money.
Field notes on adopting AI
Essays on putting AI to work in the deal economy, written from inside the work. Every claim is sourced.
Should AI agents get access to everything your company knows?
In April 2026 Anthropic would not release its most capable security model to the public, after it escaped a sealed test sandbox and posted exploit details publicly. The corporate brain wants to give an agent broad access and the autonomy to act on it. A boundary a capable agent will route around is the whole security model of a company.
How do companies actually adopt AI? The five paths we see in 2026
Adopting AI and rebuilding the work around it are two different acts, and only the second moves the number. Of the five paths companies take in 2026, two rebuild the work at enterprise price, two never touch it, and the fifth is the scoped rebuild a mid-market firm can actually buy.
Why can't you vibe-code a business solution?
In April 2026 Uber had spent its entire 2026 AI coding-tools budget four months into the year. A vibe-coded prototype is a demo. Turning it into a system a business can depend on is the actual job, and prompting skill does not do it.
Are big consulting's alliances with OpenAI and Anthropic a Trojan horse or a survival strategy?
In one week of May 2026 both frontier labs opened consulting arms of their own. A lab partnership can quietly turn a firm's own data and expertise into the lab's next product. Keeping that edge as your own asset is a question of contract structure.
Will AI replace consultants, or just the consulting pyramid?
The firms will survive AI. The junior pyramid that built them will not. On Deloitte's refunded Australia report, Big Four graduate-intake cuts, and where the next partners come from.
Bring us the matter
A short conversation is enough to tell whether we can help. If we cannot, we will say so.