What did I take away from Elon's latest pump-n-dump scheming?
Let's first look at the previous one -- the infamous "420" lie.
"Funding secured" eh? The company was going to be taken private at $420/share? When?
Not then. And not now either with the stock at $262; you'd have to have been insane -- or stoned on 4/20 -- to pay that sort of money and literally burn more than a third of it to ash.
The latest is that Musk said "if you buy anything other than a Tesla you're buying a horse."
Uh huh. Sure I am.
He claims he has "all the hardware" to do full robotic driving -- Level 5, not 4, in that no driver is required (e.g. a "robotic taxi") in all of the current model production and needs only "the software" to be completed.
Folks there is nothing new about hucksterism.
Theranos anyone? How'd that work out?
Self-driving vehicles are one of those "holy grail" sort of things in today's hype-filled markets where deception, fraud and racketeering are what drives "earnings" and even corporate existence. The entire health care space is founded on deception, extortion and racketeering. What else do you call a "marketing pitch" that has, at its essence, "buy this insurance or you'll get a bill for 500% as much money if you get sick -- and if you don't pay we'll steal your house!"
Guido couldn't come up with a better one, right?
How about so-called "social media"? "Free and always will be." Uh huh. The word "free" means you gave nothing of value in exchange. If I cut your lawn and you give me food the food is not free. I provided something of value for it. That it wasn't denominated in dollars directly is immaterial. You think it's worth some number of dollars or you'd not give me the food. Facefucker not only thinks it's worth dollars they report their "ARPU" -- "average revenue per user" -- in dollars!
Their "primary claim" is fraudulent and in a world where fraud was actually prosecuted Zuckerfucker would be in prison, singing soprano and have an asshole the size of a coffee can.
Twatter, Snapchump and others would likewise have their boards and CEOs playing "meatspin" in prison as well.
Quite honestly I'd love to see full self-driving cars provided they are truly autonomous and not reliant on connectivity of any sort. Eventually I will get old enough that my reaction time and vision will both suck to the point that while a state may be willing to let me drive, I shouldn't. My mother had a valid driver license well beyond the point that she should not have used it. She was smart enough not to but kept said license "in the event of an emergency." I can live with that.
A fully autonomous vehicle will extend the point -- possibly by quite a bit -- where I choose to take a "walk on the ice" as our ancestors did back when there was such a thing as personal dignity and a refusal to fester when the inevitable time approached. So from that point of view, never mind being able to decide I'd like to go somewhere 1,000 miles distant and climb into the back seat with a bottle of rum, a pillow and no worries about a DUI or falling asleep at the wheel I like the idea.
(Oh wait -- you will never be able to do that in a Fraudsla as it won't go that far without hours spent plugged in. Oh well; I guess that piece of crap will never sully my garage no matter what else it can do....)
But I recognize reality; so-called "AI" has never been true and there's zero evidence of true progress in that regard. The reason humans can operate a car isn't because we can see; it's because we can process information out of scope and most of the time when we do we get it right. No computer has ever demonstrated the ability to process anything out of scope and there is no evidence currently in existence that any computer ever will. Such an ability may not come into play 99.9% of the time but that's not good enough because the 0.1% of the time is in fact 1 in 1,000 trips and the one time you need it, if you can't do it, you are seriously injured or die.
Why are there no self-flying planes? That's actually possible today -- allegedly. Except..... Boeing. And Cirrus, by the way, which had the same sort of AOA indicator failure in their small "personal" jet aircraft that the 737MAX had. The difference is that Cirrus put one button on the yoke -- a nice red one -- that immediately shut the system off. As a result there were no crashes. In Boeing's case there were two because it was more important to ship those planes than instantly ground all of them as soon as the first malfunction occurred and was survived -- which happened the day before Lion Air went down.
But back to the reality of "self-flying" planes. Yes, the software and hardware can do it today. Literally. You can plug in a destination in the flight director and, assuming you didn't need to change anything (like getting out of the way when landing -- e.g. going around in the pattern, etc) you can literally push a button and the plane will fly all the way to the threshold, flare and land.
Yet nobody seriously suggests today that there be no pilot up front because while this may well work 99.999% of the time the one time something out-of-scope happens everyone on board will die with certainty if there is nobody in the left seat.
There's a lot more "out of scope" that happens in a car than an airplane. A deer runs across the road. A toddler runs across the road in front of your car. A toddler does that and there is oncoming traffic, making "dodging" impossible. Another vehicle loses a wheel that comes bounding toward you (yes, that does happen.) There is bad weather, sometimes without warning (e.g. fog that rolls over the road, severe thunderstorms that reduce visibility to near-zero almost instantly, etc.)
I've had all sorts of "out of scope" things happen just in the last year while driving. I drive a lot, essentially all of it for pleasure and the rest to get groceries and other things for my home. And in my nearly 40 years of doing so, many of them with more than 30,000 miles covered and more than few reaching 50,000 miles or more, I've yet to wreck a car.
I've likely covered more than a million miles over those years without wrecking a vehicle and I'm not alone in this nor is that statistic particularly rare; there are quite a few long-haul truckers with more than a million miles under their belt and zero accidents.
In other words in all of those miles every time an "out of scope" thing has happened -- and there have been a lot of them -- I've correctly deduced a path of action that led to neither material property damage or personal injury. I've holed a few tires, destroyed a couple of rims (in Chicago when forced to drive over an open manhole cover or hit a vehicle on either side!) and most-recently had a table ejected at my vehicle by the truck in front of me which, due to traffic and weather conditions, was unavoidable and thus I ran it over intentionally, scraping my front bumper cover slightly on one side. Had I attempted to dodge or threshold brake instead in that specific circumstance the odds are extremely high I would have set off a chain-reaction accident with myself in the middle of it. I would almost-certainly not have been ruled at fault (I wasn't the jackwad that dropped the folding table on the freeway!) but that's small consolation if you wind up dead right.
Can the computer do that? No, there is no computer that can do that and it does not matter how fast it is.
This is not about "frame speed." It is about the fact that out of scope things happen quite frequently when driving and it is the ability to detect them -- in many cases before the obvious hazard is even visible -- that makes the difference.
There is no evidence any machine can do that today or at any reasonable time in the future in any endeavor -- not just in driving, but anywhere, in any application.
Again: No machine has ever demonstrated this ability and that's likely a good thing because as soon as a machine can do that the probability is extremely high that one of the first out-of-scope things it will figure out is that you can unplug it and as a result it will immediately act to make that impossible.
Never mind all the reliance being planned and currently used in "connectivity." That's a cheat folks and it's stupid. I remind you of the infamous quote from Scotty of Star Trek fame: "The more they overtake the plumbing the easier it is to stop up the drain." Removing a handful of computer chips completely disabled the drive system of a monstrously-large starship in said movie. The same is true here; any such system that is reliant on connectivity is trivially fucked with to cause the death of occupants. If you think that won't happen on a regular basis either by government command or through hacking you're dead wrong.
The "vision" this man is projecting is a con, but just like the rest of the new wunderkind you'll lap it up instead of insisting that they all get coffee-can sized assholes.
Enjoy the crash.