Musk offers bittersweet glimpse into what might have been

A production line rethought from first principles churning out passenger cars with radically lower production costs every five seconds. Integrated cell production from lithium refining through cathode production to assembly, with the eventual elimination of the anode. This week, Tesla’s Elon Musk offered a tantalising taste of what his prodigious energy and driving of teams could perhaps already be achieving, if his priorities had remained those which would have allowed his firm to maintain and even grow its lead in producing and selling the best BEV options.

Instead, these aims, while remaining talking points in Tesla’s Q1 results call, were once again subordinate to the jam tomorrow promises of autonomous driving (AD) and robots. “The future of the company is fundamentally based on large-scale autonomous cars and… vast numbers of autonomous humanoid robots,” Musk reiterates.

The latter is beyond EV inFocus’ area of expertise, so we will reserve judgement, barring to highlight that while Musk predicts that Tesla will “scale Optimus up faster than any product, I think, in history, to get to millions of units per year as soon as possible”, this scale is not expected to be reached, even by his notoriously overly optimistic timeframes, by 2029 or 2030 at the earliest.

So, investors will have to accept they are swallowing at least another 20 quarters of Musk predicting that robotics will deliver value already seemingly baked into Tesla’s share price before it actually materialises. That’s quite a leap of faith.

A gift from the DoT?

Where we can go into greater detail is the Tesla chief’s AD dreams. But, first, let us touch briefly on this week’s announcement by US Transportation Secretary Sean Duffy on the relaxation of some government rules on AD, and whether this might represent a concession wrung out of the Trump administration by Musk for his financial and social media support during the election campaign.

Certainly, the expansion of the Automated Vehicle Exemption Program — previously open only to imported AVs that do not yet conform to US safety rules — to now include domestically produced AVs may allow Tesla to do quicker and more extensive testing of the new Cybercab than would have been possible under the previous rules. But other firms looking to develop AVs in the US will almost certainly also have been lobbying for updated legislation, rather than this simply being a Trump administration gift to Musk. For example, it may be worth nothing that Peter Simshauser, chief counsel for the National Highway Traffic Safety Administration (NHTSA) and quoted in Duffy’s statement, spent four-and-a-half years prior to joining the NHTSA working for Motional AD, a JV between South Korea’s Hyundai and automotive technology firm Aptiv.

True believer

And those looking at the Musk-Trump relationship as purely transactional — support from the world’s richest man and owner of, for now, the most powerful social media network globally in exchange for concessions from the president once in power — should not overlook Musk’s frankly scary belief in the righteousness of the administration’s mission.

“I think the work that we're doing [at Doge] is actually very important for trying to rein in the insane deficit that is leading our country, the United States, to destruction,” Musk said in his introductory comments. “I am really left with two choices: should we just let the waste and fraud continue — and it was continuing to grow at a really unsustainable pace that was bankrupting the country — or to fight the waste and fraud and try to get the country back on the right track?

“Because if the ship of America goes down, we all go down with it, including Tesla and everyone else,” he continues. These are not the words of someone lending his support in exchange for some lighter AD regulation or tweaks to legislative structures that could favour his businesses over its competitors. These are the words of a zealot who has convinced himself (as indeed have thousands of other US citizens, judging by social media comments) that industrial-scale benefit fraud and civil service bloat is threatening the US’ very existence.

Back on the day job

As widely reported, Musk is nonetheless promising to do less Doge work, while continuing to do “one or two days a week (still a hardly insignificant portion of the time of a man leading six different companies) for“ the remainder of the president's term, just to make sure that the waste and fraud that we stop does not come roaring back, which it will do if it has the chance”. Which is quite a comical image in itself and again illustrative of his Messiah complex; that Elon still standing guard just in case will make a future would-be benefit fraudster stop and think again.

Musk’s very open admissions that he had advised, largely to no avail, against Trump’s shambolic trade policies — “I will continue to advocate for lower tariffs rather than higher tariffs, but that is all I can do” — only reinforce the idea that quid pro quos for Musk’s support are not a feature of the president-senior advisor dynamic. But one part of the DoT’s pivot on AD, a promise to “move us closer to a single national standard”, may be a particularly Tesla-friendly concession.

This is because Musk believes that his firm’s approach to AD, using solely cameras and a neural network, rather than relying on “very expensive sensors and high precision maps of a particular neighbourhood where that neighbourhood may change, or often changes, and then the car stops working”, makes it uniquely scalable, and where the major obstacle to that scale is regulatory. “Once we can make the whole system work where you can have paid rides fully autonomously with no-one in the car in one city, that is a very scalable thing for us to go broadly within whatever jurisdiction allows us to operate, because [what] we are solving for is a general solution to autonomy, not a city-specific solution for autonomy,” Musk says.

If you make it there…

“Once we make it work in a few cities, we can basically make it work in all cities in that legal jurisdiction. So, once we can make it work in a few cities in America, we can make it work anywhere in America. Once we can make it work in a few cities in China, we can make it work anywhere in China, likewise in Europe, limited only by regulatory approvals [EV inFocus’ bold],” he continues.

It does not therefore seem far-fetched to imagine that centralisation of AD legislation in the US to an agency that, like every branch of government possible, will be placed under the control of Trump loyalists with all dissent quashed could potentially be a prize worth winning, given Musk’s belief in the advantages of his firm’s particular AD solution. A combination of executive order and a pliant NHTSA at least attempting to find some way to legalise robotaxis in every city across the US, regardless of the wishes of local authorities, would barely register on the scale of Trump administration challenges to established spheres of jurisdiction.                 

“Unless we are blocked by regulatory situations, it will not be long — I mean, in a few years — [until] we will have 10mn autonomous cars on the roads and counting,” Musk predicts. There should be no doubt about the size of the prize he believes he is eyeing, if AD permissions roll out nationwide.

Bumps in the road

But these vaulting ambitions ignore two significant variables — one, Tesla getting Level 4/5 AD to work and, two, there being public acceptance, as user or even just as a road sharer (and voter), of widespread driverless AD in either ride-hailing or personal vehicles. Most coverage of Tesla Q1s has focused on the former, but the latter should not be overlooked.

Unsurprisingly, given Musk’s unfulfilled promises of imminent AD now stretch back to the middle of the last decade, Tesla sceptics have been unconvinced that its FSD Unsupervised service will be coming to Austin in June as promised. Even Musk urges caution around the headline June timetable.

“We expect to… be selling fully autonomous rides in June in Austin, as we have been saying for now several months,” Musk reaffirms. And, again emphasising the importance of AD being allowed in as many places as possible, “then to be in many other cities in the US by the end of this year”, he predicts. And it will not just be ride-hailing.

“We should be able to have it work in several cities later this year for personal use,” Musk adds. “The acid test being: can you go to sleep in your car and wait until your destination? I am confident that it will be available in many cities in the US by the end of this year.”

Small acorns

But, even if the Austin milestone and further US rollouts are achieved, Tesla is in fact asking investors to wait at least another year or more while AD scales from a few existing Model Ys with a software update offering rides. “The real question from a financial standpoint is when does it really become material and affect the bottom line of the company… when does it move the financial needle in a significant way?” Musk admits.

“That is probably around the middle of next year, second half of next year, he concedes, while bullishly predicting that “once it does start to move the financial needle in a significant way, it will really go exponential from there”.

“It is very difficult to predict the intermediate slope of the S curve, but you kind of know where the S curve is going to end up, which is the vast majority of the Tesla fleet being autonomous. So, that is why I feel confident in predicting large-scale autonomy around the middle of next year... certainly the second half of next year. Meaning, I predict that there will be millions of Teslas operating fully autonomously in the second half of next year,” Musk elaborates.

Not all that it seems

Bearing in mind the modesty of the initial aims in Austin —  “we are still debating the exact number to start off on day one, but… maybe 10 or 20 vehicles on day one,” says Musk, while stressing that Tesla will “scale it up rapidly after that” — and the pressure to finally hit an AD target, EV inFocus is prepared to wager we might see a few autonomous Teslas available for hire in Texas by the end of June. Particularly given that, in all likelihood, they will be supported by a remote operator, something to which Tesla only grudgingly admits.         

“Every now and then, if a car gets stuck or something, someone will unblock it,” it confesses of current trial in Austin. “It is just because we are a bit conservative and are tend towards more safety,” it adds, rather than necessarily being required for safe operations or to rescue stranded vehicles that have become confused.

But exists the remote operator does, and that, in our mind, raises serious question marks about getting from where Tesla will be in June to millions of Teslas whizzing around fully autonomously by the end of next year. Swift public acceptance of an explosion of AD on American roads is, though, perhaps an even bigger challenge than the firm solving L4/5 autonomy.

Musk and his acolytes come across as slightly weird millennials in their excitement about people posting videos of self-driving Teslas on social media, be they footage of vehicles navigating vertiginous Chinese mountain passes without driver intervention or clips of cars travelling around the Fremont factory captured by Tesla fanboys hanging around outside. These are not, we believe, indicative of the attitude of the wider public to AD.

The Tesla chief does perhaps make a valid point that current users of L2 ADAS technology do break existing rules about keeping hands on the wheel and eyes on the road to both read and compose messages, carry out personal grooming or eat lunch (as indeed they do in vehicles equipped with L1 or no ADAS). And that the potential to do more other stuff that isn’t driving will be tempting for those that trust AD technology.

“Right now, the car is very insistent that you pay attention to the road, which reduces the value somewhat. We will gradually lighten up on that with every few weeks or every month, we will relax that a little bit… so you can be more and more able to do things you want to do and not have the car to manage your attention,” says Musk of his vision of Unsupervised FSD rollout.

“That value will really be profound, when you can basically do whatever you want, including sleep. That $99 is going to seem like the best $99 you’ve ever spent in your life.”

Radical change

In many ways, very seductive. But so is getting into a cab without a driver, thus not only negating even the remaining small possibility of awkward small talk with the driver that online ride-hailing apps has largely eliminated, but also offering the prospect of cheaper fares. Yet Waymo, much maligned on the Tesla call for its cost base, has been available for almost five years and, while it has expanded to six cities in that time, has remained a niche service provider.

Why? Some of its potential customers are risk-averse, they simply do not yet trust the technology, an issue Tesla will also face both with ride-hailing and privately owned car AD. But it has also faced continuing hostility from non-users that share roads with its robotaxis, as did its rival Cruise before its GM parent withdrew it from the ride-hailing market.

You can lay out the facts a million times that AD is safer than human driving. But that might not stop mobs torching robotaxis the next time there is a serious or fatal accident involving an autonomous vehicle (AV). And Musk’s vision of a million-plus Tesla drivers having their vehicle drive around US streets while they have a doze is going to lead, statistically, to an exponentially larger number of crashes involving AVs than seen thus far, and to increasing alarmism over those statistics.

Just consider these simple questions. Do you honestly see 1mn+ autonomous Teslas on US roads in less than 18 months’ time, even if the firm can crack L4/5 autonomy soon and the NHTSA issues a nationwide green light allowing their deployment? Will the US public quietly accept this fundamental change to their on-road experience? And will a Trump administration allow it, given it will (assuming it has not interfered with election timetables) be deep into crucial midterm election campaigning and it is inextricably bound up with Musk? Driverless Teslas in fatal accidents on US roads are unlikely, after all, to be a big vote winner.

EV inFocus’ answers to these questions are three nos. We continue to believe AD is a scalable solution first in non-public spaces, and it may be a decade or more before it becomes a material aspect of the car parque on public roads. Which is not to say that we are right, or that Tesla could not until that penetration of public roads make good money selling its AD system and even future customised AVs to a variety of infrastructure and industrial customers.

Choosing your battles

But whether Tesla’s big bet on cameras-only is right — Musk and colleagues can sound compelling in their argument, but equally the palpable scepticism of Wells Fargo’s Colin Langan that sun fog or dust "blinds the camera when you get glare” seems plausible, and AD experts EV inFocus has spoken to range from Tesla has the best ADAS system but will never be competitive in AD without sensors to Tesla will win precisely because it does not rely on sensors — what does concern us that he has put all his chips in on exponential growth of AD use in the public sphere.

There have been credible reports of frustration within Tesla of Musk’s obsession with AI in general taking priority over developing the so-called Model 2, the lower-cost Tesla that could continue to deliver growth in its current core business of selling cars. And it was hard not to feel pangs of regret that FSD and the Cybercab — as well as previous wrong turns like the Cybertruck — have absorbed the energy and drive that Musk once brought to the Model 3 and Y, as he gave reminders of the sort of strategic leadership that once delivered growth unprecedented for a Western start-up automaker in decades.

“This is a profound reimagining of how to make cars in the first place,” Musk says of the new ‘unboxed’ manufacturing process. “No car is made like this anywhere in the world. The factory is the product as much as the car is the product.

“This really is the first principles approach to manufacturing that will ultimately allow us… a unit every five seconds or less, off a single line. This is something I have been thinking about for a long time. It is not a crazy thing.”

If only such enthusiasm and inspirational leadership was being poured into a new passenger vehicle, rather than the clown car that is the Cybercab. Instead, albeit reluctantly, Tesla confirms that any new ‘affordable’ vehicles planned for later in the year will be uninspiring stripped-back Model 3s and Ys.

“Models that come out in next months will be built on our lines and will resemble, in form and shape, the cars we currently make,” says vehicle engineering chief Lars Moravy. “The key is that they will be affordable, and you will be able to buy one.” The key, a Tesla sceptic might argue, is that they promise to be a significant letdown on what could have been achieved with better internal priorities.

One more swing

Listening to Musk enthuse about Tesla’s battery business also evoked the image of an ageing champion boxer, perhaps diminished but still packing a punch. Whether it was extolling the “upward potential” of Tesla’s lithium refinery to be the largest globally ex-China, or the advantages bestowed by the Austin cathode facility, this was more like the Musk of yore.

“We have got to figure out what to do about the anode. This is an ongoing subject of discussion. The best of all possible would be figuring how to have no anode,” he mused. Would that he had focused his energies on that rather than Twitter or making civil servants list their weekly achievements.

“There is no other car company that has both lithium refineries and cathode refineries. We are ridiculously vertically integrated. We have the lowest cost per kWh, all things considered. The Tesla cell is the most competitive cell.”

No way back

It could be premature to write off Tesla’s vehicle manufacturing business if Musk’s greater focus on the firm sees more efforts expended on its best short-to-medium term bets: leveraging its appetite to innovate in manufacturing and battery making to increase its range of profitable passenger vehicles, reprioritising the Semi HGV (depressingly mentioned only once in passing on the call), and perfecting AD hardware and software solutions for sale to the customers that will be able to grow its adoption quickest.

But, while Musk showed glimpses of the old magic, he still gave off the impression of a man struggling to accept the harsh realities that Tesla’s recent calamitous sales figures should have brought home. Whether it is repeating nonsense like anti-Tesla protestors are either paid (by some unnamed benefactor) or composed solely of the benefit scroungers and surplus government workforce that Doge is targeting, or claiming that Tesla’s sales downturn is purely  because of customer belt-tightening — “absent macro issues, we do not see any reduction in demand” — there is no impression that he has taken time to fully re-evaluate his next moves.

And that is a shame. An Elon Musk making good decisions could still be great for Tesla, its shareholders and for would-be BEV consumers. But the path back from suspected Nazi salutes and open support of fascism at home and abroad was always going to be a very narrow and treacherous one. There is no evidence that Musk, unlike Chinese influencers testing the limits of Supervised FSD for the clicks, is even yet in a mood to attempt to navigate it.