When AI knows what we want

AI is becoming increasingly good at anticipating what we might want.

It can suggest what to watch, what to read, what to buy, where to go, what to listen to and even what we might want to say. Sometimes it feels helpful. Sometimes it can feel a little strange.

I’m curious about what happens to our attention when technology starts making choices before we've made them ourselves.

When something is already waiting

We don't always have to go looking anymore. Something is recommended for us. A song starts playing. A video appears. An answer is suggested before we've finished asking the question. Our attention can be drawn towards it almost without effort.

We might be interested. We might simply follow.

When we don't have to choose

There can be something helpful about having good suggestions. They can save us time and help us discover things we might not have found ourselves.

Yet when something is always being chosen for us, we may have fewer moments when we stop and ask what we actually want.

We can move from one suggestion to the next without really noticing that we haven't made a choice.

When our attention is anticipated

AI doesn't need us to tell it everything. It can learn from what we've searched for, watched, listened to or asked before. It can use that information to anticipate what might interest us next. The more it knows, the more relevant its suggestions can become. The more relevant they are, the easier it is to follow them.

Perhaps that's part of the pull.

What do we want?

Sometimes we know exactly what we want.

Other times, we discover what we want by looking around, getting curious or simply allowing our attention to wander a little. If technology is always anticipating our next choice, perhaps we can sometimes miss those moments of discovery.

We might pause before following the suggestion. What do I actually want to look at? What am I curious about? Where was my attention before this appeared?

When AI knows what we want

AI may become increasingly good at anticipating us.

Perhaps that's useful.

Perhaps it's also worth noticing what happens when something else becomes very good at deciding what deserves our attention. We don't have to ignore the suggestions.

We might simply notice them, and choose whether to follow.

I wonder what you might notice the next time AI seems to know what you want before you do.