Colorado men try to crash into a self-driving Tesla to see how it reacts and people are saying the car ended up 'annoyed'

These guys took a Tesla to a racetrack in Colorado and then repeatedly tried ‘crashing’ it on FSD.

The title of the video that Dirty Tesla and Out of Spec Reviews used includes the word ‘crash’, and that’s accurate, but only technically.

Obviously, they didn’t literally crash the EV into a wall or something, but they tried forcing it to hit soft targets strategically placed around the racetrack.

Like cones and soft pylons, for example.

DISCOVER SBX CARS: The global premium car auction platform powered by Supercar Blondie

What exactly were they doing?

They used inflatable obstacles so that they wouldn’t actually damage the car if the system failed.

But the system didn’t fail.

They tried crashing the car into these obstacles but it didn’t work.

Partly because FSD was better than expected, but also because they didn’t give it the chance to fail.

As in, a couple of times they got really close to the soft pylons but then decided to avoid them manually.

Maybe FSD would’ve failed, maybe it wouldn’t have.

We’ll never find out.

And then the car got ‘angry’.

What they mean they said their Tesla got ‘annoyed’

These days some people might be fully prepared to hear a car literally talk.

Some people might be ready to believe the Tesla they were using told them to ‘stop messing around’.

And the car sort of did that, in a way.

But it did by using tech, not words.

Like all AI systems, FSD has the ability to ‘learn’.

After several failed attempts to make FSD ‘crash’ the car, the system realized something was wrong and did two things.

First, it began prompting drivers to take action a lot more frequently and a lot quicker.

They call it the ‘nag’, which is when the car tells you (usually with an annoying sound) to nudge the steering wheel to let it know you’re still there and awake.

Then, their Tesla simply refuse to re-engage FSD.

Clever.