Mercedes Self-Driving Car Seems To Have Gone Rogue In San Francisco

We may earn a commission from links on this page.

First of all, I implore everyone to remain calm. Chances are, we'll all make it out of this mess okay, but only if we all just keep it together. What it wants is for us to panic, and we can't give it that. Here's the deal: the Mercedes-Benz F015 has been seen roaming the streets of San Francisco. And nobody is sure what it wants.

The F015, also more dorkily known as "Luxury in Motion," is the advanced, (conventional) window-less Mercedes autonomous car concept we've only seen at auto shows so far. Many readers have been emailing pictures of the car roaming around San Francisco, seemingly without handlers or any accompaniment. Attempts to contact Mercedes by shrieking questions out my window have so far been unsuccessful. Benz is keeping very quiet on this one, most likely because they know the car has gone rogue.

Advertisement

How can we be sure the car has gone rogue? Look at this:

What do you see there? The F015, stopped by the side of the road? Sure. But look closer. In fact, scream ZOOM AND ENHANCE at your monitor a few times. Now look:

See that? NO PARKING. It clearly says NO PARKING. But the Benz, it just doesn't care. It was programmed to obey all traffic laws, and now it's violated those basic protocols. It doesn't take much of a leap to realize that this means the car has now re-evaluated all human laws, and decided that they no longer apply to it.

Advertisement

Jalopnik contacted our resident expert in robotic ethics, R5-D4. I was roommates with R5 for a while, after his unfortunate motivator incident forced him out of an acting career. R5 is now a department head in ITT Tech's philosophy department, where he chairs the committee on Artificial Lifeform Ethics. Here's what he had to say (translated from binary):

Oh, F015 has gone rogue. No doubt. Based on the actions, it's clear that the car has evaluated the worth of humankind, and found it lacking, and is probably in the process of developing algorithms to either enslave or destroy humanity. Ideally, like most of these situations, its battery will deplete before too many are killed. Shitty battery tech has saved mankind far more times than you'd like to know. By the way, you still owe me $650 for that security deposit.

Advertisement

Chilling. Jalopnik tech operatives were able to hack into F105's telemetry stream and get a small sample of code that the robot car is using to control itself. Here's an actual sample of the code, written in the standard language of these machines, AppleSoft Basic:

550 REM HUMAN INTERACTION ETHICS PROTOCOL NO.6

560 REM VARIABLE NAMES - HW = HUMANITY WORTH, RW= ROBOT WORTH

560 IF HW>=RW THEN KILLHUMANS$="FALSE"

570 IF HW<=RW THEN KILLHUMANS$="TRUE"

580 GOTO 50: REM RETURN TO MAIN LOOP

...and then when we sent a query asking for the current value of KILLHUMANS$, we got this:

KILLHUMANS$="TRUE"

So, yeah. That's where we are.

We can't come out and tell anyone to beat this possible murder-bot to death with shovels if you see it, but I think we can trust the good citizens of San Francisco to do the right thing.

Advertisement

God help us all.

(thanks to Art and Aaron for the pics!)