Principal

I recently noticed that it's a bit clunky to talk about "users" of our agents. Building complex enterprise software, we have multiple types of users, from operations teams making daily decisions to executives looking at high level dashboards, and of course suppliers. But often none of these people can make important judgement calls, like "will this cheaper deal with supplier B be better than the existing deal with supplier A, even though B's items are not a 1:1 match".

So when we say "our agents act on behalf of X", it'd be great to have a generic word for X without always specifying the persona. Fortunately this is a solved problem: it is the principal.

I connected the dots through the principal-agent problem but I tried to trace it towards its roots. This is Wikipedia:

In commercial law, a principal is a person, legal or natural, who authorizes an agent to act to create one or more legal relationships with a third party.

This seems to go all the way back to Roman law, with "principal" becoming a noun later in medieval civil law.

As far as I know, today software (like AI agents) cannot enter legal relationships on your behalf. This limits the applicability of agents, in addition to capability and security problems of course. But I expect we'll get there.

In any case, the concept of a principal is very useful in thinking about how agents will be adopted.