Should companies that develop AI have a "VP of AI ethics"? Here is a downside of titles of this sort, and an alternative.
➤ Downside: Signals of passing the buck on value judgments
Responsible AI involves deciding on difficult ethical questions.
For example, suppose a company wants to prevent bias in its dataset. Doing that requires deciding which social groups should be protected from bias, to what extent, who to prioritize in cases of conflicts, and so on. These are, in part, value judgments.
In other areas of business, management is often in charge of making the tough calls. For example, the VP, or chief, of finance will make tough calls about the budget. The VP, or chief, of technology will make the tough calls about R&D.
By analogy, a title like "VP of AI ethics" can trigger expectations that this person will make the difficult choices about ethical questions.
However, in my opinion, when it comes to value judgments we all have to make up our own minds. Like other value judgments, such as who to vote for, we cannot outsource these decisions or assign them to others.
➤ Alternative: "Responsible AI Advocate"
It is helpful to have someone who is leading the efforts to develop AI responsibly. An alternative model is to see that person as in charge of facilitating conversations about value questions with the goal of reaching communal decisions. They contribute professional expertise, but they don't force their own value judgments on others.
Additional important responsibilities may include building responsible AI strategies and helping to implement them, making sure things get done, raising awareness, educating, and so on.
A title that makes sense to me for this kind of work is "Responsible AI Advocate". It's not a perfect title, but I think it's a step in a good direction.
➤ Read the LinkedIn discussion here