The dishonesty of The Dawn Project’s anti-Tesla Super Bowl ad
An advocacy group called “The Dawn Project” aired an ad challenging the safety of Tesla’s Full Self Driving feature during the Super Bowl. The ad is a good example of dishonest fearmongering.
Before I analyze the ad, let me clarify two things.
First, this ad appeared, not nationally, but in local markets like Washington, DC and Atlanta. (Local stations get a proportion of all network TV advertising, which is why you saw ads for car dealers and roofers along with the expensive beer and carmaker ads in last night’s Super Bowl). Local Super Bowl ads don’t cost nearly as much as the $7 million per spot that national advertisers pay.
And second, I agree with the premise of this ad. I am a Tesla owner. I don’t trust Tesla’s autopilot, which is the limited, default driving system you get if you don’t pay extra. I didn’t pay the $15,000 extra for Tesla’s “Full Self Driving (FSD)” feature, but from everything I’ve seen and read, it is flawed enough and makes enough mistakes that I would never trust it. Tesla’s sensors miss things and Tesla’s AI makes mistakes — and in a moving vehicle, mistakes are dangerous.
But the issue of the fairness of The Dawn Project’s ad is independent on where you stand on the safety of Tesla FSD.
What’s wrong with the ad
Here’s the 30-second spot.
Shocking stuff. Clearly, Tesla Full Self Driving isn’t perfect. But I have questions.
First, how many times did the people filming this ad set up these situations, and how many different settings did they need to try, to film the Tesla doing these shocking things?
Did it fail every try, once every 10 tries, once every 100 tries, or once every 1,000 tries? That makes a difference. A human driver might miss these things once every 1,000 tries, but wouldn’t fail on every try. Is FSD better than a human?
And second, when the ad says “90% agree” that FSD is endangering the public and should should be banned immediately, 90% of whom? 90% of the Dawn Project’s contributors? 90% of people shown this ad? 90% of a random sample of people pulled off the street? (Unlikely.) The 90% figure is meaningless unless we know the sample and how the question was asked.
This ad is dishonest.
Full Self Driving is a beta test happening with all the drivers on the road
Yes, NHTSA (the National Highway Traffic and Safety Administration) should regulate FSD. It should do more tests and report data we can trust.
And yes, Tesla’s CEO Elon Musk has shown himself to be a loathsome and prejudiced individual who often makes up “facts” about his companies — including a staged demo of Full Self Driving.
But if the only way you can challenge lies is to be deceptive yourself, nobody benefits.
Let’s test Tesla. But there’s no need to get down in the mud yourself to do that.
The Dawn Project is only an advocacy group on paper. For future information, Dan O’Dowd is the founder of The Dawn Project. If you actually go through their website the ‘Dawn Methodology’ that they tout as needing to be an industry standard and being the core of their campaigns can’t actually be found. Sure there’s a bunch of flowery words about how “The Dawn Methodology enables people to write software that never fails and can’t be hacked” and how “Without The Dawn Methodology, debugging is like fighting bats in a cave” and how “The Dawn Project personnel are the Special Forces of software.” Which is odd, considering how few advocacy groups produce things.
They give several prominent examples of systems developed using ‘The Dawn Methodology’. The B1-B, the B-2, the B52, the 787, a handful of F-series fighters, and the A380, oddly specific examples. Remember Dan O’Dowd? Founder of The Dawn Project? Turns out he has a day job at Green Hills Software, a company that, oddly enough, worked on notable projects like the B1-B, the B-2, the 787, a handful of F-series fighters, and the A380, as the President and CEO. Guess which market Green Hills Software has been trying to break into lately: Full Self Driving.
Take that information as you will, but I’d recommend a reach extender because something smells fishy.