Elon Musk is a polarizing figure. His ideas frequently come about in casual conversations. People are often amused and impressed by his achievements. I must admit, a few years back I thought he is literally the next Steve Jobs, only actually better, since he was onto so many things... I admired SpaceX, thought that Tesla cars had many great solutions in them...
At some point in 2015 or 2016 Elon started talking outrageous stuff in the domain of AI, a domain of my own expertise, which I could tell right away was total bullshit. And then I began looking at all this stuff in detail. Doing some math here and there. Reading various opinions. As a result, my opinion on Musk and many of his ideas has changed somewhat substantially. At this point, I can pretty much say with confidence that 90% of his stuff is utter BS, and the remaining 10% is perhaps impressive but still questionable.
Nevertheless he is quite a character with many fans almost religiously believing everything he says. Any time I meet somebody who is a Musk fan I have to go over these issues so I decided to write this post as a point of reference. If you consider yourself a religious believer in Musk, you will be offended by what follows, you've been warned. I don't care. We can talk in a few years and see who was right.
This is arguably one of the most developed of the ideas and it almost looks like it is a success. SpaceX has now been landing boosters almost routinely and a few of these landed boosters went back once or twice on other missions. However... there are still a few outstanding questions.
First of all, let's get that behind, vertically landing rockets is not a new idea. NASA landed astronauts on the moon decades ago with a rocket. In the 90's another project DC-X sponsored by the military shown a successful demo before funding got chopped. No one in the engineering community ever questioned whether it would be technically feasible to land a rocket booster on a barge.
What people did question and what remains to be seen, is whether such model is actually cheaper than using disposable rockets. To see why this question is far from being answered one has to remember that the space shuttle program had exactly the same aim: a reusable space plane with reusable boosters. Only the main fuel tank burned up in the atmosphere but that was not a huge sacrifice, since the tank did not have engines, which are by far the most expensive part. However, although good looking on paper, it turned out to be a commercial disaster. The shuttle had to be essentially taken apart after every mission (and even then there were a few close calls and two full blown disasters). Ultimately lifting cargo on a shuttle was much more expensive than using a regular, disposable rocket.
The rocket as it performs ascent and then descent undergoes some serious accelerations (way higher than a typical plane), while being built extremely delicately to minimize the mass (much tighter margins than an airliner). The rocket engine undergoes much higher stress than a jet engine. Hence all that stress can cause material fatigue and small malfunctions, while the margin for error on a rocket is extremely small. Hence if e.g. it turns out that the risk of mission failure on these reused boosters increases substantially after every use, all the anticipated savings from reusability may be quickly wiped out by one or two lost missions. And it should be noted, that the cargo that goes into space is often MUCH more expensive than the rocket itself.
So essentially we are exploring here the amount of tail risk that we can shave to get some savings. And for that we need data. A lot of data. And the fact that someone did it once or even ten times is not sufficient to declare a victory. Each of the space shuttles flew dozens of times, when one of the SpaceX boosters will have flown 100 times with between flights maintenance at an economical level, then it will be an actual success. And we are not close to that yet.
And before anyone starts rambling, I do find SpaceX launches inspiring and I take my children and show them the rocket trail whenever one is launched from California. But as much as I would love to see SpaceX (and other such endeavors such as Blue Origin) succeed, the jury is still out on whether this makes sense at all. We just need more data.
Flying from NY to Shanghai in 30min
This marks the first of Elon's ideas that makes no sense at all. In fact it is completely atrocious. First thing is that as presented the rockets would be launching and landing close to city centers. That is completely out of question for a number of reasons. One of which is that a rocket is a tank full of high explosives, something like a 20KT bomb. If anything were to go wrong, a city could be demolished by an explosion [see the Halifax explosion to get an idea of what that would be like]. Second, it would most certainly take a lot more than 30 min to get to any place on Earth. At the orbital velocity it takes approximately 90 minutes to do a full circle, so the lower limit to get to the other side of the planet is ~45min. And this is at orbital velocity. Assuming optimistically that the ascent and decent would take ~8min each, we end up with approximately 1h. Given that the rocket would have to be at least 5 miles from any urban area, the passengers would likely need to be transported via a helicopter into the launch area. And these would not be your regular passengers, since not everybody would be healthy enough to take 2.5-3G acceleration on both way up and down (likely more on the way down). The cost of such travel even with full reusability would easily exceed that of a private jet travel, and above everything, it would be vastly more dangerous. Anyway, all things together, this is pure daydreaming without any connection to reality.
SpaceX uses an animation on their webcast of a transformation of Mars into a green planet. I typically view that as a joke, even if we knew how to do it, it would take centuries if not millennia to accomplish, fair enough. But what is more seriously discussed is to send humans to Mars by say 2030 and have them set up a permanent structure with the aim of colonization.
This sounds very romantic, and I think that if we stretched ourselves quite a bit, we could probably send an Apollo style mission to Mars, mainly to plant a flag and get some samples. That alone would be dangerous as hell and really technically difficult, but permanent structure, Mars city, colonization, these are just sci-fi musings.
Mars is a very hostile environment for humans. The atmospheric pressure is 1% of that on Earth, for all practical purposes it is vacuum. The temperature rarely goes above freezing and often falls below the condensation temperature of carbon dioxide. There is almost no magnetic field, hence there is substantial amount of cosmic ray radiation. The distance to Earth is large, in case of emergency there is no way to evacuate (journey from Mars to Earth takes at least 6 months with optimal orbital alignment, otherwise it would take substantially more than a year). And above everything, there is really no incentive to go there. If there was a mountain of gold or some other highly sought material, but frankly it is like Arizona desert-land only at -100C, in vacuum and 60 million miles away. The only thing that speaks for Mars is the abundance of carbon dioxide and water which given enough energy could be used to produce rocket fuel.
The argument often brought up here is that humans need to go multi-planetary to avoid extinction. I'm willing to accept that argument, but the disaster necessary to make Earth less hospitable than Mars would have to be truly epic. If we detonated all the nukes ever made, we would not get even close. There would still be atmosphere, plenty of water, manageable temperatures. It would seriously need to be a planetary scale cataclysm, way beyond the thing that killed the dinosaurs 65 million years ago. So this argument as it stands is bullshit.
Personally, I'd love to see more space exploration, as in general I think this is the way to go for our species in the long run. But I'd go in small steps, first focus on the Moon. First of all because we have already been there and we know what to expect. Secondly, a trip to Moon takes 3 days and it's relatively straightforward to evacuate. And third, there is plenty solar energy available there, total lack of atmosphere makes landing easy (Martian atmosphere, although thin, makes spacecraft heat up on descent). We should just find some cave on the Moon and start building a permanent base there. Perhaps deeper under Moon's surface we would find the necessary resources such as water and carbon dioxide (water seems to be present in the always shaded craters). If we learned how to kickstart manufacturing there, the Moon would be a great outpost for launching more distant missions further down the road. Aiming at colonizing Mars now is simply silly.
To conclude, as of the end of 2018 SpaceX hasn't flown a single human to orbit. Neither have they launched any interplanetary mission. They did ship a Tesla roadster into the asteroid belt but apparently missed an orbit. And we are expecting this same company to launch humans to Mars in 10 years from now. For the perspective, Apollo program took 9 years from inception to moon landing at the cost of roughly 100 billion of todays dollars. I may be wrong, but I'm 97% confident we won't see any human on Mars by 2030.
Vactrain or Goddard train is not a new idea. It has been around for more than 100 years, and over all those years nothing ever came out of it. There are several reasons why this idea only looks good on paper: it is outrageously expensive and superbly fragile. Let's start with the first: maintaining vacuum in a large structure poses quite a few challenges. The atmosphere at sea level presses with a force of ~1kg per square centimeter. That is 10 Tons per square meter! In order to withstand such constant inward force structures have to be extremely strong. To understand what kind of forces we are talking about here, take a look at this video of a train car tanker imploding. The largest vacuum chamber currently maintained is the Large Hadron Collider, and I don't have to argue here that it is extremely expensive. The structure is underground in a protected tunnel under constant supervision and maintenance. To build any meaningful hyperloop, one would have to build an extremely solid pipe, take the air out (which itself would cost a fortune) and pray that nothing happens that could compromise the structure at any point (such as e.g. the mundane problem of thermal expansion). Any small structural issue would rapidly propagate catastrophically and destroy the entire tube. Now we are asking to put inside capsules moving at ~300m/s. It is again, a dream. There is no way that would be economically and technically feasible and we did not even get to how these capsules would be boarded, how would the maintenance be performed etc.
And frankly we have airplanes... They travel at ~300m/s, in thinner atmosphere (pressure at cruising altitude is roughly 1/3 of that at the sea level). Luckily 11km above the ground there are rarely any solid bodies that could penetrate the hull, which being a pressurized cylinder gets extra mechanical strength. And finally, all one needs to land a plane is a strip of flat pavement. This is exactly the reason why Goddard Train aka"hyper-loop" idea was dead for the past 100+ years while aviation flourished beyond anything the Wright brothers could have imagined, and my bet is that this will remain the case for the foreseeable future.
Not much is known about this endeavor aside from some bold statement that make all the neuroscientists I know roll their eyes.
This relatively recent idea appears to be some sort of light version of hyperloop, no vacuum and a car inside. This is advertised as a solution to mass transit. Now again, this is simply ridiculous.
First of all, electric cars in tunnels carrying people en masse is an old idea. It is called a metro, and it has been known and used all over the world for more than 100 years. Now let's dig to some numbers: a typical metro car weights some 40 tons and is able to take 175 people (probably more if they squeeze). It means, the infrastructure mass is some 228kg per passenger. A Tesla model X weighs 2.7 tons and takes 5 (max 7) passengers. That is the mass of the infrastructure is 380-540kg per passenger. Right there, without even going any further into other things, the raw cost of mass of infrastructure carried, which translates to the amount of energy necessary to operate is roughly 2x that of actual mass transit. A train does not need to carry a battery for example, just takes the energy from the third rail. In addition a train has much lower rolling friction etc...
This is just a tip of the iceberg. There is so much wrong with this idea that it is hard to know where to start, this blog post goes into many many other issues that render this idea a complete delusion. Also Thunderf00t has a nice debunking video with the analysis of the alleged cost savings of drilling Elon bragged about.
I've written several articles , , ,  on this blog about the self driving car hubris and the general AI hype. Autopilot as impressive as it is as a cruise control system, is not by any means autonomous driving ready. Anyone who thinks otherwise is an idiot, period, and should be rapidly taken out of the road because he is endangering innocent people. The fact that this technology has been released as "beta" to general public is irresponsible and significant number of fatalities already occurred.
Solar shingles, lithium batteries for home
Elon Musk unveiled solar roof tiles in October 2016. Since then some prototype installation have been rumored, but the product has not been opened to the general public. Why so? It would seem like they figured it out no? Well, again like with many Elon's promises the devil is in the mundane engineering details. The shingles themselves are certainly doable, and probably there are ways to make them pretty. The problem is electric connectivity. Each tile needs to be connected to others, these connections have to be element resistant (it is a roof after all!) and reliable. Getting that to work is not easy and fragile. Wind, tiny earthquakes or even birds walking along the roof displace the tiles ever so slightly and they loose contact. Once the circuit is broken it is difficult to find and fix the fault. Even worse a short circuit can lead to heating and fire danger. Turns out regular solar panels are way easier and cheaper to work with.
Aside from that, Tesla encourages people to buy a giant Lithium battery for home, the so called powerwall. Again this looks nice and all that, you generate energy, store it and then charge your electric car with it - a sustainable future fairy tale from a guy who regularly commutes with the largest private jet available. And again, when actual math is done the results are somewhat different. Dave Jones from EEV Blog has a very comprehensive summary of the Solar Panel math and economics. His analysis is for Sydney Australia, but many characteristics there are actually quite similar to southern California, LA is latitude 33 degrees North, while Sydney is 33 degrees South and has very similar weather pattern. In short - a large costly battery is simply uneconomical. If you desperately need to have a protection against grid failures, it is a much better idea to buy a small UPS with an auto-on diesel generator (at a small fraction of the cost of the battery). Unlike the battery, the generator could be refueled and serve as backup power for days in case of a prolonged grid failure such as in case of an earthquake or other natural disaster. Several KWh battery will only provide juice for a few hours and costs an order of magnitude more, while it will not be able to pay for itself over the expected lifetime.
Shenanigans at Tesla
I have to confess, there are a lot of things about Tesla cars I like. They look fresh, their design is clearly distinguishable from other brands and they look elegant. The minimalistic design of the interior is reminiscent of the Star Trek TNG enterprise and looks good. Perhaps they went a few steps too far with Model 3 lack of cluster display. I think Model 3 with a HUD display would be really neat, but just the center screen and lack of many physical knobs is too much for me (and potentially dangerous). And yes, the electric drive with instant torque is a unique experience (unique to EV, not to Tesla though).
Now that the prerequisites are behind let's get to the main point: while every other car manufacturer on this planet simply makes cars without bragging about it, Tesla is a never ending drama... Constant scaling up, ramping up, getting manufacturing right... And constantly it turns out their new manufacturing recipe, the alien dreadnought ends up being a tent in which workers crowd between mountains of boxes and manually assemble the cars loaded onto the line by a forklift - a far cry from modern car manufacturing titans such as Toyota. Tesla has issues with stuff long solved by every other manufacturer: panel alignment, leaking seals, paint job blemishes, rattling plastics, all sorts of QA things. And I'm not making this stuff up, twitter, Tesla forums and even fan based vlogs on youtube are full of complaints and horror service stories. This may have been growing pain for a 4-6 year old niche car company, but Tesla has been around for more than 15 years, burned through tens of billions of dollars [mostly debt which Tesla has roughly $10B of] and is valued on the stock market above Ford and GMC. These sorts of issues, the Model 3 bait and switch (the $35k so far does not exist) are unprecedented. I won't even go into the mountain of issues the short sellers bring up - even if only one third of their findings are true, this company is in a complete management chaos. What is certainly true is the long list of executives departing one by one over the last few years, this is indicative of authoritarian management and toxic company culture to say the least. So all in all, certainly not everything is right about this company, and I'm sure all these short sellers are there for good reason, they smell bullshit.
And finally there is the 420 buyout fiasco which lead the Securities and Exchange Commission to charge Musk with 20mln dollar fine and ban him from being a chairman of the board for 3 years. Later in an interview Elon stated that "it was worth it", and that he "does not have the respect for SEC". At this point he just behaved like a giant a$$hole, there is nothing to add.
Cave rescue fiasco
In summer 2018 wold's attention was on a group of children stranded in a cave in Thailand. Elon in his Tony Stark manner offered a rescue with a submarine. Vern Unsworth, a diver involved in the rescue operation expressed his opinion on the sub idea in a famous interview. Musk responded with calling this guy a pedophile, after which he apologized but then doubled down by saying that since the guy did not sue him it must be true. When Unsworth rolled out a law suit Musk's lawyers responded to dismiss the action stating essentially that Elon talks shit on twitter anyways and this should not be taken seriously.
I'm not a cave diver, but I did dive in a few small caves and wrecks and I can fully understand Unsworth argument. Orientation underwater is limited, there is a lot of gear to maintain, and passing this torpedo-tube through a tight cave would be dangerous, very slow and ultimately pointless (the rescue ended up succesful without the use of magic submarine).
Secondly, the whole pedo thing and the lawsuit are not something an adult successful entrepreneur would get himself into unless he felt some deep insecurities. To me it just exposes Elon as an arrogant and narcissistic buffoon, not some capitalist superhero.
VTOL electric supersonic jet
Another one of these ideas out of nowhere, this one I actually did some calculation myself to see how feasible it is. Here it is: boeing 737 has a takeoff weight of 72 tones and needs ~96600MJ to get to 35,000 feet cruising (just for the climb). Assuming a battery of say 25 tones, that would require 3.8MJ/kg specific energy, that is for climb only (and only 35,000 feet). I'd add another 2-4MJ/Kg for the actual cruise. So given same aerodynamic efficiency as B737 we are talking roughly specific energy of at least 8MJ/kg (with 25T of battery). Also as the typical jet burns the fuel it gets lighter, electric plane does not. The energy spent on climb will effectively be recovered in the form of velocity at descent. Anyway, we are talking 8MJ/kg, almost twice as much specific energy as TNT (and 6x the density of any chemical battery in existence). If we'd like to go supersonic, that would effectively double the energy requirements (need to climb much higher and overcome a much higher drag during cruise). If we are talking VTOL that adds a quite a bit to the climb but above everything, makes the design much more complex and heavy. So overall we are somewhere in 15-20MJ/kg range in the optimistic scenario. At 15MJ/kg we are talking about an extremely powerful explosion if anything were to go wrong with the battery. Not to mention that no chemical battery at that specific energy exists at the moment. One can tweak some of these numbers one way or another but in my view there is no way to make this work. Skipping many other details, one can firmly conclude that this idea is ridiculous.
Although resistance to Elon and his fans is futile, and we will all be assimilated, I call bullshit. I think the crowds of people who think Elon is the savior of man kind will be in for a great disappointment. Extravagant entrepreneur, great salesman, dreamer he is. But second incarnation of Jesus, savior of the planet and genius engineer he is not. Time will tell if I was right.
If you found an error, highlight it and press Shift + Enter or click here to inform us.