Warren Redlich @WR4NYGov is one of Elon Musk’s most fervent disciples. He said in a video on March 29 2022: “I think that Elon Musk is the greatest person to ever walk this earth”! Recently Warren published a vicious YouTube video containing a torrent of false accusations and invective about my intelligence, character, and motives. No doubt because I offended his faith by criticizing Elon Musk’s FSD! Let’s analyze this product of his intellect and character.
Of course, Warren needed to start off with a bang, so he nicknamed me Dirty Dan. Is that the best nickname he could come up with? Was choosing this nickname intended to persuade others of his credibility, maturity, and fairness?
He claims that I am a Tesla hater.
This is a ridiculous lie. I don’t hate Tesla. I have said repeatedly that I own three Teslas. One of my Teslas is featured in my short film “My COVID Year” which is available on YouTube. I have not driven anything but a Tesla for over 11 years. I have no short position, puts, or any other financial interest in Tesla. I am trying to save the successful Tesla electric car company from its association with the disastrously ill-conceived Full Self-Driving software that is nowhere near reliable or secure enough to be delivered to 60,000 customers (and growing rapidly), let alone tens of millions of customers in coming years.
He also said I was “Bat shit crazy”, an “absolute fucking moron”, “bleep”, and “bleep”.
His mother should wash his mouth out with soap! Does this use of profanity provide any additional information to his audience about the veracity of his claims? Or does it just convey his growing desperation in his inability to defend his opinions against the growing torrent of evidence that I am putting forward?
“doesn’t care about protecting human drivers”
Another preposterous lie! I point out safety defects in a software product that thousands of lives depend on and he claims I don’t care about protecting human drivers on our roads. I care very much about the safety of the public and believe that this unsafe software should be banned from our roads immediately.
“Dan O’Dowd wants more people to die.”
This lie plumbs the depths of slander! I thought he was a lawyer. He has a fool for a client. Here Warren descends to Omar Qazi’s level (see here). I have dedicated my life and hundreds of millions of dollars to figuring out how to make computers safe for humanity (dawnproject.com). Why does he propose that we turn our children’s safety over to a defective AI that doesn’t know that it is wrong to crash a car into a pedestrian, cyclist, or another vehicle? Is it because that is required to justify the preposterous stock valuation theories that he shamelessly promotes on his YouTube videos, where Tesla stock will go to $177,000 per share and he will become filthy rich?
“The evil Dan O’Dowd”
I am going to call this a lie. Ordinarily, I would just sluff this off as a meaningless emotional outburst, but from the extremely emotional tone of his voice it seems like he has lost touch with reality and he really means it!
I dedicate 80 hours a week to saving humanity from people like Elon Musk and him by blocking and reversing the blind rush to trust millions of lives to software, like Full Self-Driving, that is riddled with bugs and security vulnerabilities. He seems willing to say or do anything to make money on his Tesla stock.
“idiot”
Another lie! How can people tell that I’m not an idiot? Firstly, by my career achievements. Secondly because when he published a nine-minute tirade of irrational rants, lies, insults, and vicious fabrications about me, I realized that I could use it to demonstrate to an unbiased audience that there is no point in listening to anything he says. It makes it obvious to everyone that he is just an emotional propagandist and shameless stock promoter who will spout any unsupported fabrications to make more money on his Tesla stock.
I can think and act rationally, even when provoked by a vicious and unfair attack. He is driven by emotion rather that facts, evidence, and logic. When anyone dares to question his faith he reacts irrationally, making it easy to discredit him.
“Dan O’Dowd is either and absolute fucking moron, or he’s a liar, or he is a big, big, big, ass liar.”
More lies. I’m not a moron. I covered that above under “idiot”. What facts, evidence, or logic was he trying to convey with his abusive language? Or was he just expressing his emotions again?
What information was he trying to convey with the addition of “big, big, big, ass” to the earlier choice of “liar”? That he is vindictive and petty and unworthy of serious consideration?
“And he has a new rant on Twitter today, a couple a tweets. And he’s lying, he’s misrepresenting, he’s deceiving and it’s really, really offensive on a day a bunch of people are mourning the death of 6 high school girls.”
That is a really, really, low attempt at character assassination! He accuses me of being “really, really, offensive” for doing something I do every day because people were mourning the sad deaths of six high school girls who died in a car accident which I had not heard about. What exactly did he not do that day because he was mourning those girls? And why is what he did that day not “really, really, offensive”?
Now we come to the substance of his video. The rest of it being just meaningless profanity and insults. He critiqued two tweets I posted highlighting safety defects in Full Self-Driving software that are demonstrated in two YouTube videos.
First Video
In this video it sure looks like FSD is trying to turn left into oncoming cars. But he asserts that his detailed frame by frame analysis of a wavy blue line demonstrates that the turning of the steering wheel to the left was done by the driver to prevent FSD from turning to the right and crashing into another car. Either way it is just another Full Self-Driving safety defect!
Second Video
He asserts that the beta test driver in this video hit the brakes too soon, and if he had not, the Tesla would not have hit the pedestrian. But he wasn’t there, the driver was. And the driver clearly says that the Tesla would not slowdown in time on its own. He insists over and over that the beta test drivers are very conscientious, alert, carefully tested and selected, right up to the moment that Full Self-Driving malfunctions and tries to kill someone. Then he says the driver screwed up. I will take the evidence from the person who was there and saw everything, over the self-serving fabrications of a Monday morning quarterback who wasn’t there.
“He’s either evil, a liar, or an idiot. Or, possibly all three. So, Dan take your pick. What are you? Are you evil. Are you an idiot? Are you lying? Or are you all of the above?”
Warren’s sophomoric multiple choice question didn’t include the correct answer: Elon Musk is wrong about Full Self-Driving and I am right. No facts, evidence, or logic seems to undermine his blind faith in Elon Musk.
In over 25 years of developing and implementing some of the most safety-critical and security-critical systems in the world (See https://dawnpr and oject.com/about-our-founder/), I have learned far more about software reliability and security than ElonMusk or anyone else at Tesla. In my expert opinion, the development processes being used for Full Self-Driving ensure that it will never achieve the reliability, safety, and security required for full self-driving. And all the facts, evidence, and logic are on my side.
“I’d love to hear from you Dan. What’s your take on this? Dan O’Dowd shut the hell up.”
Which one is it then? He doesn’t really want to hear from me. He wants me to “shut the hell up”. He really is irrational. He should try thinking before speaking.
“I know from FSD’s development that this will save thousands of lives in the United States and millions of lives worldwide.”
“It’s an open question whether we’ll get there or not.”
Again, which one is it? Either he “knows” it will save millions of lives or “it is an open question whether we’ll get there or not”? The contradictions that tumble from his mouth have no information value. He can’t “know” it will save millions of lives based on facts, evidence, or logic because he isn’t an expert in AI technology, software reliability, or software security. He can only “know” it will save millions of lives through his blind faith in Elon Musk.
“We’re getting really close.”
His opinion that Full Self-Driving is really close to attaining full autonomous driving is not based on facts, evidence or logic. He even said “it was an open question”. And yet Waymo self-driving software has reported over 30,000 miles between disengagements, while FSD is lucky to go 10 miles between disengagements in urban driving.
Every year for the last eight years Elon Musk has promised full autonomous driving within a year or so (see video here). He was wrong eight times. The most likely conclusions based on facts, evidence, and logic is not that he will reach full autonomy this year (the evidence is 8 to 0 against that), but rather that next year he will say that it will be available next year (the evidence is 8 to 0 in favor of that).
In an urban environment, Full Self-Driving malfunctions more often than any other product, automotive or not, currently being sold by any respectable company. Full Self-Driving is the worst software on the market. The problem with Full Self-Driving is that it is stupid: after years of “training” and thousands of instances of negative feedback from beta test drivers, Full Self-Driving still hasn’t learned the most basic lesson, that it is wrong to crash into people, cyclists, and other vehicles. It malfunctions more than 1,000 times more often than the average human driver (see our research on dawnproject.com). Full Self-Driving doesn’t know right from wrong, and no one understands its decision process as to whose life it is going to save and who it is going to let die. It would take a lot of faith to entrust the lives of our children to an unstable AI. Indeed, we deserve to become extinct if we do!
After eight years of failed software development, four failed hardware systems, and eight failed directors of the autonomous driving program at Tesla, what has been produced is the worst commercial software on the market. It can’t even learn the most basic principles of driving. Facts, evidence, and logic would conclude that Full Self-Driving will never even be close to the reliability required for fully autonomous car software.
Millions of lives will depend on autonomous driving software, so it must be the best software ever written. Full Self-Driving has been developed in accordance with Silicon Valley’s “Move Fast and Break Things” philosophy. This always results in bug-ridden products like Full Self-Driving. Facts, evidence, and logic would dictate that Full Self-Driving will never approach the reliability of the best software ever written.
After eight years of nothing but failure, Elon Musk has fabricated a fantasy that the world’s stupidest AI, the worst software on the market, will suddenly, miraculously, and exponentially accelerate its pace of learning to reduce its malfunction rate to become a thousand times more reliable by the end of the year. There are no facts, evidence, or logic to support this fantasy. No amount of training can fix plain stupid. It never gets easier and easier to fix bugs when reliability increases, it always gets harder and harder to fix bugs as reliability increases. This fantasy is the last desperate hope that after eight years of nothing but failure, a miracle driven by faith can save Full Self-Driving.
The entire factual basis for Warren’s torrent of lies, accusations, insults, and slander, is:
• a difference of opinion as to whether a split-second of the first video shows that Full Self-Driving turned the steering wheel to left to try to crash into another car, or whether it turned the steering wheel right to try to crash into another car; and
• a difference of opinion between him and the driver of the car in the second video as to whether the Tesla would have stopped in time to avoid hitting the pedestrian;
Neither difference of opinion justifies his torrent of invective, or anything more than a gentlemanly agreement to disagree.
I understand the frustration that causes him to lash out. His faith in Elon Musk has been sorely tested again and again by the repeated failures of Elon’s preposterous promises about full autonomy. I know how angry he is that I keep pointing out facts, evidence, and logic that show that Full Self-Driving will never be reliable or secure enough for widespread deployment, and that it will never save millions of lives. He gets even angrier when other people believe me. And even angrier as I turn out to be right again and again. Yet he clings to his blind faith in Elon’s FSD.
Warren should abandon the false god that has made him so angry. He would then feel a lot better. He should look at the facts rationally, not filter them through his emotions and reject them due to his blind faith.
There is a culture that puts safety and security ahead of everything else, including functionality and time-to-market. This culture must pervade all aspects of autonomous driving software development. We must define the rules for safe and secure operation of an autonomous vehicle, and we must demonstrate that the software always follows those rules. This will be the best software ever written. It will be designed, written, documented, reviewed, tested, validated, and certified to the highest standards of software safety and security. These processes have worked excellently for over 25 years to produce many autonomous computer-controlled systems that are far more complex than a car, and that have never failed and never been hacked.
If it takes a long time and costs a lot of money to create demonstrably safe and secure autonomous cars, then so be it. If it turns out that we don’t have the technology needed to build fully autonomous self-driving cars in a way that we are certain is safe and can’t be hacked, then we should not build them.
The Full Self-Driving development process based on Silicon Valley’s Move Fast and Break Things methodology of delivering defect-ridden partial functionality to customers, with the promise of making it safe and secure later always results in a final product that is riddled with defects and is easily hacked. That is acceptable when the worst thing that can happen is that you need to reboot your computer. But it would be a threat to humanity if we were to bet billions of lives on its safety and security, when we know it is not safe or secure.
It is imperative that we make computers safe for humanity.
Nothing less is acceptable.