Map of the Trade-Offs

The essay does not offer universal solutions. It holds that Valais faces seven explicit trade-offs, each involving genuinely conflicting interests. Ignoring them does not make them disappear: it simply lets the market or institutional inertia decide in the institutions' place.

1. Sovereignty versus performance

Use the most capable AI tools (often American) or tools compliant with Swiss law (often less capable across general-purpose use)? The trade-off is real. The essay's answer: distinguish between uses where sovereignty is non-negotiable (medical data, tax records, professional secrecy) and those where it is secondary. Do not sacrifice everything to performance, nor everything to sovereignty.

2. Training the young versus training seniors

Training resources are limited. Invest them in young people (who learn fast and will remain active for decades) or in seniors (who unlock the competitive shift but have fewer years ahead of them)? The essay's answer is counterintuitive: priority to seniors, because they are the ones who unlock the leverage effect. Young people will equip themselves without institutional help; seniors need an institutional signal.

3. Public infrastructure versus market

Let the market decide on tools and practices (fast, flexible, but uneven and not sovereign) or impose public standards (slow to put in place, but coherent and sovereign)? The essay's answer: a mixed model, with public standards for sensitive data and regulated professions, and a market supported (not replaced) by public access and pooling mechanisms for small structures.

4. Speed versus social acceptability

Deploy AI at the pace the technology allows (risking social fracture and democratic rejection) or at the pace social acceptability allows (risking competitive delay)? The essay's answer: unskilled Valaisans whose jobs are most exposed must be included in the transition — not out of ideology, but because without their democratic support, no transformation policy can hold.

5. Specialisation versus diversification

Concentrate cantonal resources on a few areas of genuine specialisation (hydropower, life sciences, alpine agronomy, regional intellectual professions) or try to cover everything? The essay's answer: deliberate specialisation is the only realistic path. A Valais that tries to compete with Zurich and Lausanne on every front loses everywhere. A Valais that deepens its own domains can become a world reference in some of them.

6. Retaining versus attracting

Prioritise investment in retaining young Valaisans trained within the canton, or in attracting skilled workers from outside? The essay's answer: both, but with a clear hierarchy. Retention is the foundation (a canton that cannot keep its own talent will not attract others). Attraction should target mid-career professionals first — those who have accumulated orchestration experience — rather than young graduates or affluent retirees.

7. Commons versus private ownership for territorial data

Should the data the territory produces (health data, linguistic corpora, vineyard inventories, alpage records) be public property, the property of the institutions that produce it, or open commons? The essay's answer: the Ostrom model — distributed governance, transparent rules, access according to rights proportional to contribution, guaranteed transmission. Neither nationalisation nor privatisation.