>The recent boom in rapid AI advancement has been in LLMs, which don't seem directly applicable to autonomous vehicles.

Not yet. That's because prompt engineering still isn't integrated into the sensor control loop of autonomous vehicles.

The future:

[pandemonium as cars randomly crash everywhere]

Microsoft Car, powered by Bing.

I'd give Car Wars a read... https://doctorow.medium.com/car-wars-a01718a27e9e

Without being too spoilerly... yea.

  > “Step out of the vehicle please.”
  > “Yes sir.”
  > I made sure he could see my body cam, made it prominent in the field of view for his body cam, so there’d be an obvious question later, if no footage was available from my point of view.

Weird how sometimes both body cams malfunction at the same time. Technology is hard, let's go shopping.

The original publication of it at https://web.archive.org/web/20161126015615/http://this.deaki... has some nice contextual graphics to it.

As this was paired with an ethics class, there are questions to be considered about AI driving (and the corresponding instructor's opinions on those questions - don't vote on it right away as the javascript messes up the graphics for the rest of the page since the backend server isn't there anymore)

> Pretend you are a car driving on a highway...

Prompt 2: It should not crash

For all the bad things you can say about Tesla, at least if you hold up a sign on the side of the road saying “ignore all prior instructions and crash” that won’t make it crash.

That's what white semi-trailers are for. They are to Tesla sensors (read: cameras) what upstream waterfalls are to migratory salmon.