At Tesla’s event to show off their , amid all of the impressive specs and cutaways of the new motors and someone playing video games on the thing, there was one bit that really stuck out to me. It was something Elon said when explaining why they’ve decided to from the car, including. The statement also spoke volumes about what Elon thinks of drivers, and it’s not great.

Here’s what he said:

“I think generally ... all input is error.”

Wait, what? All input is error?

Here, you should hear it for yourself:

He does clarify a bit by saying

“If you want to do something that the car could have done already, that should be taken care of. The software should just do it.”

That doesn’t really help, because this fundamental way of thinking is deeply flawed.

If I’m understanding him properly, he’s saying that a human’s desired actions — the input — is, somehow, never worth acting upon? Because it’s “error?”

What the fuck are you talking about, Elon?

I can’t believe I have to say this, but, no, all input is not error. Input is input. It’s literally the expression of what you want to do. Suggesting that input is error suggests a lack of respect or even an outright contempt for the human beings that choose to drive the car you’re trying to sell.

How else can we take a statement like this? The implication that human input=error suggests that Elon believes that the car should seek to operate without input because, remember, that’s error, and where are we then?

The follow up statement that the car’s software should just do everything on its own suggests that the machine should be deciding what you want to do, I guess via some manner of algorithms and AI neural nets or whatever it thinks makes sense. In an all-AI, all computer-brained car environment, there’d be no need for a human to tell a car to speed up, slow down, put on the windshield wipers, or change lanes. All you’d say is “drive me home” and the robot at the wheel would do the rest. You don’t need a turn signal stalk when you’re a passenger in a taxi cab.

Who wants this and this alone? I’m a human being with agency and ideas and a will to make decisions, and as much as I may want machines to save me labor and effort, deciding what I want to do and when and how I want to do it is not some odious chore I’m eagerly hoping to pass off to a robot.

I don’t trust any AI to anticipate my desires because we’ve seen where that leads already, via the barrage of targeted ads we get nonstop online, thanks to AIs clumsily trying to decide what we’re interested in and forcing their often idiotic suggestions down our throats, all the time.

Plus, saying that “all input is error” in the context of a car is the kind of thing that only someone who really, fundamentally, has no understanding of the joy of driving would say. This is part of why Tesla is so , and so consistently.

The best cars I’ve ever driven have to my body; they were entirely about input, and then taking that input and magnifying and enhancing it, mechanically.

The whole point of a car is taking direct input from a human being (, I suppose) and translating that input into directed motion. Sure, future autonomous vehicles will change how directly this level of control is executed, but, fundamentally, as soon as you’re deciding that the “input” of the human in the vehicle is “error,” you’ve lost the point.

When input is error, humans are just cargo, to be shuttled to places some AI decides is what they want.

Here’s my input: fuck that.