AI博覧会 in 博多: 5 sessions, one shared problem

The writer went in with a practical goal: as an instructor, to find inspiration that could solve problems inside their own company, and to pick up better ways of using tools for day-to-day work improvement. What they found instead was a pattern repeated across sessions. Departments not coordinating, and know-how concentrating in individual people, came up again and again — something the writer connects to earlier experience at a job-change event their company exhibited at, and to a coffee-industry event called SCAJ from their barista days. The writer treats the two problems as linked: when information stays inside one department, others never notice the opportunity, and when knowledge stays with one person, it is hard to pass on.

The proposed fix, as the writer frames it, is not a specific product but a capability. Companies that already know how to use AI can build their own tools in-house and find solutions themselves; companies that don't end up using tools made by those that do. The writer warns this can cost more than paying directly for something like Gemini, Claude or Codex, and comes with limited customization plus the effort of getting it accepted inside the company — including the persuasion, and possibly the disagreement, that brings.

The stakes therefore seem to hinge less on which AI service a company picks than on whether it can develop solutions for itself. The writer's case rests on one attendee's impressions across about five sessions rather than any formal survey, so whether this pattern holds more broadly is likely something readers would need to test against their own workplace.