Questions worth thinking about.

There is something seductive about capability. Give people a new tool and the first question is usually what it can do. This is reasonable. It is also incomplete. Technology increasingly allows us to ask “can we?” The more important question is “should we?” The answer is rarely available in the product demonstration.

Building begins before the building. It begins with choosing a problem, deciding whose experience matters, and noticing what an apparently elegant solution might make harder. A system can be efficient by its own measures and exhausting for the person required to use it. Another login is not, in itself, a contribution to civilization.

I keep returning to ordinary decisions. What are we eating? What do we already have? What matters to the people around the table? These questions are small only from a distance. Up close, they contain time, money, health, culture and care. The point of food technology is not to make dinner resemble an administrative procedure. Dinner has enough stakeholders already.

Purpose should come before the instrument. If the purpose is to help people make a better decision, then more options are not automatically better. Neither is more information. Sometimes useful work means reducing the burden: bringing the relevant things together, making the trade-offs understandable, and leaving the person with genuine agency.

That requires restraint. We should be able to explain why something belongs, not merely why it was possible to add it. We should also be willing to remove it when its demands exceed its value. A feature can be impressive and still be an unwelcome guest at the table.

Worthiness is not a promise that every decision will be right. It is a discipline for making decisions and accepting responsibility when they are not. Who is this for? What problem actually matters? What deserves to survive? Are we making life better—or just adding another login? Those questions do not produce a particularly dramatic launch video. They may, however, produce something worth keeping. That seems a better ambition than being briefly impossible to ignore.

Artificial intelligence can process more information than any person ever could. That does not make judgment obsolete. It makes judgment more important. An answer can arrive quickly, fluently and with remarkable confidence. None of those qualities tells us whether we asked the right question. Confidence has never been especially difficult to manufacture.

Intelligence is useful for identifying patterns, comparing possibilities and organizing complexity. Wisdom asks what matters in this particular situation, to these particular people, with these particular consequences. The distinction is not a complaint about technology. It is a request that we give technology the right job rather than quietly promoting it beyond its qualifications.

A decision is not simply an information problem. It contains priorities, constraints and values. At the family table, a meal might fit a preference but ignore what is in the kitchen. It might suit the pantry but not the people eating. A useful recommendation has to make sense in context. The longest list is not necessarily the most helpful answer.

The same question applies to institutions. Information can clarify a choice without making the choice for us. Someone still has to decide which trade-offs are acceptable and take responsibility for their effects. Delegating the analysis does not delegate the obligation. A dashboard is not a conscience, even when its charts are exceptionally attractive.

This is why agency matters. Good technology should help people understand their options, not make disagreement feel like user error. It should make room for uncertainty and correction. It should know when a recommendation needs a human decision rather than another decimal place. More precision is not always more understanding.

If information becomes abundant, what becomes scarce? Attention, context, restraint and practical judgment are reasonable candidates. We should design for those things deliberately. The question is not whether machines should become less capable. It is whether their capability helps us become more thoughtful, more responsible and less overwhelmed. Technology is the instrument. Better decisions are the work. Wisdom remains the willingness to ask what should actually happen next—and occasionally to conclude that nothing needs to be automated today.

A founder who remains permanently indispensable may have built a successful company without yet building a durable institution. This is an uncomfortable distinction. Indispensability feels flattering. It also means that a considerable number of ordinary decisions may be waiting for one person's reply. That is not quite the same thing as leadership.

Founders bring energy, conviction and a willingness to begin before everything is clear. Those qualities matter. But the responsibility changes as the work grows. The question becomes less about what the founder can personally carry and more about what the organization can understand, decide and sustain without constant interpretation from its original author.

Stewardship begins with purpose. Why does this work exist? Who is it meant to serve? Which principles should remain recognizable when the product, team or circumstances change? If the answers live only in the founder's head, the institution has a memory problem. A shared purpose needs to be understandable enough to guide decisions, not merely attractive enough to print on a wall.

It also needs room for judgment. Building a durable organization is not the same as writing instructions for every possible Tuesday. People need context, responsibility and permission to challenge an assumption. Governance should make good decisions possible without making every decision a ceremony. I say this with the affection and caution of a recovering bureaucrat.

Preservation offers a useful way to think about the obligation. Cities have memories. So do institutions. What deserves to survive is not always the most visible thing. Sometimes it is a standard of care, a difficult lesson, a relationship or a reason for saying no. Stewardship means taking responsibility for those things without insisting that everything remain exactly as it was.

Ownership asks what is mine. Stewardship asks what has been entrusted to me, and what condition I will leave it in. The two need not be enemies. But they are not interchangeable. A founder's ambition can include building something that outgrows the founder's centrality. Serve. Preserve. Build. The form of the work changes; the obligation to use it well does not. Being necessary at the beginning is understandable. Making yourself necessary forever is a rather demanding succession plan.