Three weeks ago, I sat down with Evan Troxel, and we covered a lot of ground:
- Why visual scripting tools don’t scale into enterprise products
- How LLMs are democratizing automation for non-coders
- The real automation opportunity: emails, reports, and operational data rather than model generation
- AI tools that interact directly with legacy software that has no API
- Building a firm-wide second brain from project emails, specs, and past decisions
- MCP as the protocol that lets AI agents talk to fragmented tools in plain English
- Why automation is a commodity and firm data is the actual IP
- Internal skill libraries as shared, reusable infrastructure for teams
Then I listened to Alain Waha at Bricks and Bytes. Same network, no coordination, entirely different altitude.
It’s interesting how we ended up covering a lot of the same topics without even talking to each other.
This kind of overlap is worth noting because it points to something bigger. When people in the same industry, looking at it from different perspectives, start landing on the same points, it’s not just a coincidence. That’s a pattern crystallising.
Take the PDF problem, for example. Every decision made during a project eventually turns into a PDF. Alain looked at it from a C-suite angle, calling it a massive collection of unstructured documents that holds the industry’s knowledge hostage. I came at it from the workflow side: when AI can actually pull some structure out of those documents, info from Project A can finally help out Project B. Our perspectives were different, but our conclusions were the same.
The cultural shift follows a similar pattern. Alain framed it from a strategic angle, saying companies need to decide if they want to be seen as a value brand or a low-cost competitor. This choice really comes down to whether they treat knowledge as something to share or something to guard. I looked at it from an operational view, emphasising the need to codify expertise into systems that everyone can use, build internal libraries, and capture what engineers know before they leave the company. Again, we framed it differently, but we diagnosed the same issues.
We landed on common points, because we are reaching points (very quickly) which are hard to ignore.
But I believe in patterns. Once you start to see similar thoughts stick across independent conversations, across different altitudes and different professional contexts, that’s an inflection point for change. It may not be the change itself, but it’s the moment when you can see the direction things are heading.