Apparently, Humane’s lapel-squatting phone-replacing everything-device doesn’t “do apps”. It instead automatically chooses “AI experiences” to interpret and act on users’ commands.

I hate that they’re called “AI experiences”, because we’re inherently passive when we experience something: it happens to us. And yet we should be inherently active when we use a device. I wish these things were called “services”, because that’s what we should expect from a device that should be a tool.

But to give a device like this the air of being a tool would be to establish with it a relationship in which we expect discerning and decisive control. To be active. And modern technology companies don’t want us to have discerning and decisive control over our devices. They want us to be passive consumers of a series of functions defined by technical and economic systems that are completely outside our control.

I hate this technological inversion of our relationship with tools, which is turning us from being active to passive.

Here are three simple predictions for Humane’s AI experiences (which kind of make a sick joke of its maker’s name):

• They will obscure the internal logic they use to interpret our needs.
• They will limit opportunity to become skilled at using them.
• They will not allow us to actively adapt and modify them.

All these principles are anathema to a good tool.