Finally, a step-by-step guide for even the skeptics to try to see what spot the LLM tools have in their workflows, without hype or magic like I vibe-coded an entire OS, and you can too!.
I also love the soundtrack so much and have listened to it thousands of times, especially By The Wall, my favorite song. PS: Thanks for posting the composer’s solo name, Floex, because there were (are?) two people with exactly the same name working at Amanita Design, bizarrely!
It's just that the company has stalled every major project they started, and, so far, completed a rather shitty an uninspiring one in Vegas that has no reason to exist in the first place (it's subway but with Teslas instead of trains).
Its only purpose is to prevent the money from being spent on viable public transportation projects, and in that sense, it's very interesting that it got so far.
Or, digging tunnels is a lot harder than expected and there have been no big technical leaps to change that. The idea is great, but only if the cost goes down and digging speed goes up by a lot.
I assume you got a cut of the $23bn my state took with the promise of a high-speed rail, which afaik is the only "viable(?!)" transportation project that could have been affected by this, or you just hate subways/subterranean transportation progress?
The idea isn’t great, tunnels for cars or pods have really low throughput (low occupancy + safety margin headways, even at a high speed). And it hinged on them magically revolutionising an already highly mature field, which surprise, surprise didn’t work out.
If it had been possible to speed up and reduce the cost of tunneling, the thing that would most make sense is running regular trains through them. But they never had any real ideas to actually make it cheaper or faster (apart for making it too small for proper emergency egress), just the idea that SV tech guys would be able to find a way to do it.
I disagree, tunnels for transport of both people and goods, especially in high-density urban areas is the best way to go. Walking and biking is great for their distance, but cars and trucks are still needed for larger and heavier items. Using shared transportation (like a train) is terrible for "The last mile". Doing everything at night just seems like a band-aid and sucks for all those workers.
The idea of trying to solve the hard infrastructure problem of digging first also seems like a great idea. Build the aqueduct before you build the millions of houses and farms, and even let anyone do that part.
It's still premature to say that they haven't revolutionized the field, people around the world are still digging tunnels so there's still a market. It wouldn't be the first time an already highly mature field got revolutionized, I still don't get why you're so anti-tunnel.
Here is the thing, you demand incredibly high amounts of capital cost for not actually achieving much. And that capital could be use for far, far, far, far more useful things the city actually needs. Like high capacity transports, like metros, trams or bike lanes.
The amount of goods that need to be transported to stores and such things isn't that big. And using literally free unused roads at night or early in the morning is just a great deal.
For individual transport last mile is regularly being done by cargo bike or small electric truck just fine.
But you are right, tunnels do make sense for some things. Like transporting garbage underground. Or transporting heat underground for district heating, or district cooling. Both would be better investments then trying to move logistics under-ground.
There is a reason, no serious attempt anywhere in the world is trying to move logistics under-ground. There are just so, so many better ways to invest in the city. Its literally not even in the Top 100 most needed things.
Specially in the US where the road network is so hilariously overbuilt that it could serve 10x the amount of people on the same area if public transport was just taken minimally serious. And in the US, underground cargo transport isn't even in the Top 1000 things a city should consider spending money on.
There’s only so much gridlock you can avoid without going above or below grade. I was shocked when I moved to Seattle and they had no subway system. It was made even worse by being crammed up against a tall hill with a ridiculously deep lake behind it. They are finally changing it now but I’d spent time in Tokyo before, and time in London and Paris shortly after and it was a real head scratcher. One bus tunnel helped, as was evidenced when they shut it down for a couple years, but cmon.
Because The Boring Company hasn't built any tunnels worth noting, perhaps, and stalled most of its projects.
This isn't a case against tunnels, this is a case against The Boring Company.
The tunnels aren't a great idea apriori. Good luck pitching the tunnels idea in Venice.
The tunnels may be a good or a bad idea depending on many variables, and the tunnels that the Boring Company has actually built are worthless.
As for the tunneling equipment: selling those machines isn't their core business, and there's no evidence these machines have, or may in the feasible future, do anything revolutionary in the tunnel industry (i.e., built radically better, radically cheaper, or radically faster).
The idea of having such machines is good. They don't have such machines.
> It's still premature to say that they haven't revolutionized the field
It's never premature to say that. Read what you wrote.
You can say a field has been revolutionized once a revolution takes place.
It's hasn't.
The impact of the Boring company on the way tunnels are dug is, very sharply, zero.
Is it possible that they will? Sure. It's also possible that Britney Spears will. She still has the time, it's premature to say she wouldn't do it, right?
Their mistake was to go from tunnels to transportation systems. I'm sure there are some innovations possible in tunnel boring. But that's not going to be some massive growth market.
> But trying to reinvent transportation was stupid.
It's not stupid, it's weaponized incompetence to divert funding from actual transportation infrastructure to their non-solutions which are all about the company owner's biggest money-making product (cars).
> It seems like the motivation here is that Musk is aligning Tesla to a narrative that justify the absurd stock price, even if that narrative isn't reality.
Since Tesla stock has always been 90% based on the narrative, the narrative is the reality (and the product) of Tesla, and the actual machinery made and sold are just props and decorations to create the impression of it.
Maybe they should rebrand themselves as poTemkin: keep the T logo and the mysterious Slavic vibe, while shedding the pretense about what they're about.
Won't affect the stock anyway. Everyone knows the company is overvalued based on promises and perception alone.
Everyone's just betting on the charade going on one moment longer than their hold on the stock.
If you squint, the Cybertruck is shaped like a pyramid on wheels, which couldn't work any better as a visual metaphor for the enterprise.
This Daily Mail article¹ has it. It.. doesn't look brutal to me?
Just looks like the minibus driver, who was driving on the median, veered across it into the oncoming lane to crash with the semi.
He wasn't in a lane to begin with.
> Allegedly caused by lane assist activating out of the blue
Yeah dawg, imma need a second opinion on this.
This is alleged by the survivors of the crush.
Which is weird, because the passengers wouldn't know about what happened in the split-second that resulted in the crash.
Particularly, the passengers wouldn't know about whether lane assist interfered.
And the driver, who would, also happened to be drunk and high AF on cannabis, cocaine, and yet-to-be-identified stuff found in the vehicle at the moment of accident⁴.
Methinks, these allegations might be a lil' biased.
* * *
EDIT: the other comment revealed the news that the vehicle did not have a lane assist feature.
An failure that didn't live up to the hype that generated the initial sales volume in pre-orders.
The idea of the Cybertruck sold well — at a time before Musk's Roman salutes, shadowing Trump, running DOGE, and further enshittifying what remains of Twitter.
The actual Cybertruck, once the pre-orders ran out... did not.
Nearly half of all Cybertrucks sold (about 75% of those sold in 2024) were pre-orders.
That's to say, people stopped buying once they saw the Cybertruck for what it actually was (ditto for Elon).
- have a body that's not literally duck taped together in some places and can easily snap in others
- use steel (which bends) for body construction
- be suitable for towing hauls
- not be ridiculously overpowered (...to the extent where engine can overpower the breaks)
- have good visibility with a windshield that isn't at a sharp angle to the ground and body geometry which doesn't maximize blind spots
- not have sharp corners that the cut you or doors that can decapitate your dog
- have door handles that make doors openable in case of emergencies/no power situations/electric shorts
- not have bulletproof glass (WTF, "for the masses"?) which makes makes it harder to rescue people when accidents happen
- be easily repairable, or at least amenable to repairs in local non-Tesla shops, with customers being confident it their warranty won't go poof (as the law requires)
- be easily customizeable for different applications (particularly when it comes to the bed)
- not look so different from other trucks without any reason other than "Elon Musk wants to be edgy": ugly is subjective, being a billionaire's fashion statement isn't
...to start. That's off the top of my head.
And, of course, being priced for the masses, which doesn't just happen. It's a design requirement.
As it stands, the Cybertruck is, and has always been, a rich boy's luxury toy — and it was designed as one.
It really seems like something got to Musk's head that he thought the world has so many edgy rich boys.
You want to see a modern truck "for the masses"? That's Toyota IMV 0, aka Hilux Champ [1]. Ticks all the above boxes.
And hits the $10,000 price point [2]. A literal order of magnitude cheaper than the Cybertruck.
Speaking of which: a car "for the masses" isn't a truck. It's a minivan (gets the entire family from one place to another), it's a small sedan/hatchback (commuter vehicle), a crossover/small SUV to throw things, kids, and dogs into without having to play 3D Tetris in hard mode.
But not a pickup truck, which is a specialized work vehicle.
The masses aren't farmers and construction workers (most people live in the cities, and only a small number needs such a work vehicle).
The popularity of The Truck in the US is, in a large part, a byproduct of regulation which gives certain exemptions to specialized work vehicles.[3]
That's not even getting to the infrastructure part: trucks shine in remote, rural areas. And while one can always have a canister of gas in the truck bed, power stations can be hard to find in the middle of the field or a remote desert highway.
But again, it's not impossible to make a truck for the masses (at least for certain markets). That's the $10K Hilux Champ.
For all the luxury aspects of the Tesla sedan, it's been one of the most (if not the most) practical electric vehicles on account of range alone. It also looked like a normal car at a time when EVs screamed "look at me, I'm so greeeeeen!" from a mile away (remember 1st gen Nissan Leaf or BMW i3?). It was conformal and utilitarian, while also being futuristic and luxurious enough for the high price point was fair for what was offered.
The public image of having a Tesla was good: you are affluent, future-forward, and caring for the environment.
The Cybertruck went back on everything that made Tesla a success: it's conspicuous, impractical, overpriced, and currently having publicity rivaling that of the recent Melania documentary.
It was not a risky bet. It was an a-priori losing bet. The world simply never needed as many edgy toys as Musk wanted to sell.
And driving a car shaped as an "I'm a Musk fanboy" banner really lost its appeal after a few Roman salutes and the dear leader's DOGE stint.
Overly optimistic engineering assessments? Perhaps, but they are much further down on the list of reasons of Cybertruck's failure.
I was talking about mass market for an electric truck.
- F150 is big
- Its perfectly usable, claim otherwise are nonsense. Arguable depending on your workload it has advantages. Not being as good for side-loading is a downside, but many people can't side-load F150 either. But having a cover that locks safely is clearly an upside. In terms of what people actually use these trucks for, like shopping or picking up a few things from Home Depot the bed is useful. Secondly, beds are empty 99% of time anyway.
- All electric trucks are not perfect at towing loads over long distances. For short distances its very good. And again 99.99% of time people are not towing loads long distances. The issue is really only if you want to tow loads long distances as fast as possible.
- Visibility is better then F150
Most of the rest is just nit-picking or looking at the issue only from one side.
And you completely ignore that F150 is already a truck for the masses, as it is literally the most sold vehicle in the US, and it doesn't have to cost 10k. Comparing the Cybertruck to something like Toyota IMV 0 makes no sense when F150 was the target.
> The popularity of The Truck in the US is, in a large part, a byproduct of regulation which gives certain exemptions to specialized work vehicles.[3]
Something that is often claimed but isn't true. That's a contributing factor but by no means the only reason.
Which is why I like this article. It's realistic in terms of describing the value-propositio of LLM-based coding assist tools (aka, AI agents).
The fact that it's underwhelming compared to the hype we see every day is a very, very good sign that it's practical.
reply