AI as infrastructure, not feature.
The office treats artificial intelligence the way the previous generation treated electrification — as the substrate every serious business, institution and community will be re-platformed against.
Human capability, multiplied.
Eby's interest in AI is not academic. It is operational: how does a clinic actually run with AI in the loop? How does a school? A mortgage broker? A municipal council? The answers — when worked out honestly — are how opportunity gets distributed at scale.
Conviction, not consensus.
Amplify, don't replace
AI should make humans more capable. Any deployment that erodes judgement, agency or dignity is a failed deployment.
Distribute, don't hoard
The economic value of AI must reach the operators, the students and the workers — not only the platforms that ship it.
Deploy with discipline
Models are tools, not strategies. Strategy is the organisational change AI makes possible.
Ethics is infrastructure
Safety, transparency and consent are first-order engineering concerns, not communications problems.
Build for the next billion
If a product only works for the technologically affluent, it isn't AI — it's a luxury good.
"The question is not whether AI changes everything. It does. The question is who gets to participate in that change, and on whose terms."