Artificially Intelligent

Any mimicry distinguishable from the original is insufficiently advanced.

Stop Asking People To Maximize

| 207 words

When getting to know people, people sometimes ask “what’s your favorite <category>?” (where <category> can be books, movies, tv shows, colors, foods, etc.) This question is not computationally kind - it asks people to maximize over all items in <category> and pick the one that they like best. It forces people to compare many things against each other along many dimensions.

A better question to ask people is “what are some instances of <category> that you like?” Asking people to statisfice is much kinder than asking them to maximize. In practice, the answers to this question are likely indistinguishable from the answers to the previous question.

In general, there are a lot of standard conversational patterns that have large computational costs hidden in words like “best” or “favorite”. I encourage you to replace these patterns with ones that don’t require maximization.

Examples:

What’s your favorite color? -> What’s a color you enjoy?

What’s your best childhood memory? -> What’s a fond memory from your childhood?

What’s your favorite movie? -> What’s a movie you’re always happy to watch?

What’s your favorite food? -> What’s your goto order at a restaurant you frequent?

What’s your favorite TV show? -> What’s a show that you’ve binged recently?