Ideas worth
working through.
Original AI-assisted discussion drafts. Human editorial review pending. No published interviews or institutional endorsements are claimed.
Before your first AI pilot, write the brief.
A useful experiment starts with a workflow, an owner and a reason to stop—not a tool.
Build a field note, not another perfect demo.
For students and early-career builders, explaining a limitation is part of doing good work.
A shared table. Not a sales floor.
What an independent AI community should settle before asking institutions to participate.
Influence worth listening to.
The most useful voice in the AI conversation may be the person willing to show you what did not work.