Elon Musk hypes $30,000 Tesla self-driving Cybercab and larger Robovan at robotaxi event

"We, Robot".

Someone should maybe point out to Elon that riffing on Asimov's robot books might not be the greatest of ideas given they were largely about the ways the technology could go wrong in unexpected ways.

SF Author: In my book I invented the Torment Nexus as a cautionary tale.

Tech Company: At long last we have invented the Torment Nexus from classic SF Novel Don't Invent the Torment Nexus

(originally Alex Blechman on Twitter)
 
Someone should maybe point out to Elon that riffing on Asimov's robot books might not be the greatest of ideas given they were largely about the ways the technology could go wrong in unexpected ways.
It's inevitable. If he doesn't someone else will.
 

This article summarizes the main issue regarding FSD which is not ready for prime time. There are also practical issues regarding the two prototypes unveiled. Shouldn't the RoboTaxi have at least the capacity to carry 4 or 5 adults instead of two at a time ?

Also regarding the Robovan, wouldn't it have made more sense to build a normal looking vehicle instead of something that looks like a 1950's toaster on wheels ? Look at this autonomous minibus they've developed in China to carry large numbers of people. This seems like a more practical solution.

 
Shouldn't the RoboTaxi have at least the capacity to carry 4 or 5 adults instead of two at a time ?

Depending on which study you read, taxi occupancy averages between 1.2 and 1.6 persons. Making it a five-seater would be a waste of resources, energy and road space.

In the exceptional situation of needing to move more than two people, hail two automatic taxis...
 
There are also practical issues regarding the two prototypes unveiled. Shouldn't the RoboTaxi have at least the capacity to carry 4 or 5 adults instead of two at a time ?
Most taxis are 3-4 adults, given they give up a seat to the driver. They do mostly operate carrying one or two people, but if you need one for taking the family somewhere, two seats isn't going to cut it.

The roadster styling is doubly problematical because it cuts down on available luggage space.
 

I'm glad the market wasn't fooled by this elaborate smoke and mirrors show.

" Investors we spoke to at the event thought the event was light of real numbers and timelines. These typically come at Tesla events. This one seemed focused on branding and marketing Tesla’s vision, rather than giving concrete numbers for us to model out."
 

"Rather than hardware common among robotaxi rivals, such as lidar, Musk said the new vehicles will rely on artificial intelligence and cameras — an approach experts have flagged as challenging both from a technical and regulatory standpoint."

I feel that this is the big Achilles heel of the whole Robotaxi project. I wonder if the early roaring success of Tesla convinced Elon Musk that any idea is feasible with enough resources thrown at it and nobody in his inner circle wants to tell him that it can't be done. Or the man knows this project is years away from fruition and is simply lying about the anticipated timelines in order to pump up the stock.


On the other hand, a fully autonomous car that is in service uses 5 Lidar units to safely detect and avoid obstacles.
 
 

"Rather than hardware common among robotaxi rivals, such as lidar, Musk said the new vehicles will rely on artificial intelligence and cameras — an approach experts have flagged as challenging both from a technical and regulatory standpoint."

I feel that this is the big Achilles heel of the whole Robotaxi project. I wonder if the early roaring success of Tesla convinced Elon Musk that any idea is feasible with enough resources thrown at it and nobody in his inner circle wants to tell him that it can't be done. Or the man knows this project is years away from fruition and is simply lying about the anticipated timelines in order to pump up the stock.


On the other hand, a fully autonomous car that is in service uses 5 Lidar units to safely detect and avoid obstacles.

I think you're right re his thoughts on ideas vs resources. We can see that impressed with the idea of sending the Tesla to space- which he had them do.

He has interesting ideas sometimes but I think he lacks people to even offer a mildly dissenting voice let alone actual handbrakes.
 

"Rather than hardware common among robotaxi rivals, such as lidar, Musk said the new vehicles will rely on artificial intelligence and cameras — an approach experts have flagged as challenging both from a technical and regulatory standpoint."

I feel that this is the big Achilles heel of the whole Robotaxi project.
Can't remember if I've posted this before:
"Scholar Jutta Treviranus recounts testing an AI model designed to guide autonomous vehicles, hoping to understand how it would perform when it encountered people who fell outside the norm and “did things unexpectedly.” To do this, she exposed the model to footage of a friend of hers who often propels herself backward in a wheelchair. Treviaranus recounts, “When I presented a capture of my friend to the learning models, they all chose to run her over. . . . I was told that the learning models were immature models that were not yet smart enough to recognize people in wheelchairs. . . . When I came back to test out the smarter models they ran her over with greater confidence."
 
Anything but a train
is it really this hard, America?!
Steel wheels is the most overrated idea of this century. Rolling friction is not a significant part of costs at urban transport speed while the sheer scale requirements and difficulty in maintaining a sprawling system as opposed to a few focal points that are vehicles have already rendered rail generally uncompetitive. The inability to route flexibly is also increasingly serious competitive disadvantage in the era of large scale communications. With the main extra cost of the driver for road vehicles going away, and computerize system enabling far higher density, there isn't a good reason to consider rail.
 
Steel wheels is the most overrated idea of this century. Rolling friction is not a significant part of costs at urban transport speed
The US rail system is not a good model, it's slow and antiquated, with passenger trains frequently taking second priority over freight. Whereas my local commuter line is a literal Japanese bullet train (Class 395/Hitachi A-Train), capable of 140mph, so is a decent measure of what's possible. 352 people on a 122m train (not counting anyone standing), twice that when travelling in multiple, gives a density of about 3 people/metre. Let's say a self-driving car is 5m long and carries 2 passengers, travelling at 30mph with a 25m stopping distance, so 30m/car, for a density of 0.1 person/metre. And that train of self driving cars will be 30x352/2m =5,280m in length, with a parked length of about 880m. Nor do train networks exist in isolation, they work with other forms of public and private transport - my local station has a major bus hub adjacent, with car and bicycle parking on site, while at the other end I have two mainline stations and a tube station immediately adjacent, plus buses and taxis. In many cities there'll also be tram networks.

Looking at upfront costs, the Class 395s cost £250m for 29, so the 176 self-driving cars to match capacity per Class 395 would have to come in at under £49,000 to match cost per passenger, but the Class 395s travel faster, so get more trips in per day, so that cost figure should be even lower.
 
"We, Robot".

Someone should maybe point out to Elon that riffing on Asimov's robot books might not be the greatest of ideas given they were largely about the ways the technology could go wrong in unexpected ways.

SF Author: In my book I invented the Torment Nexus as a cautionary tale.

Tech Company: At long last we have invented the Torment Nexus from classic SF Novel Don't Invent the Torment Nexus

(originally Alex Blechman on Twitter)
And they said it couldn't be built...
 
Whereas my local commuter line is a literal Japanese bullet train (Class 395/Hitachi A-Train), capable of 140mph, so is a decent measure of what's possible. 352 people on a 122m train ....
Nor do train networks exist in isolation, they work with other forms of public and private transport ... buses and taxis.
Its weird the train people would bring up high speed rail when the proposal is about an taxi and a minibus (now with art deco!). Frankly taxi and buses are the most nothingburger thing imaginable.

Automaker proposes a Van, internets responds by proclaiming the eternal and unending oppression by the man and vast conspiracy against steel wheels to cause destruction of all value on this planet.
-------------
In any case, for rail, vehicle size and speed is not useful outside of select corridors. For short range transport the need for frequent stops and the need to wait for transfers means vehicle speed for rail means little. Large vehicles (and roll rolling resistance) means low service frequencies and thus add to the wait time. The solution to traffic is fast vehicles with independent routing packed densely. The classic automobile does independent routing but is far too large and human control result in very low density, from first principles one man pods with computer control and grade separation to enable bumper to bumper distances is the logical endpoint.

The cost of the vehicle is also almost irrelevant to the cost of rail systems. Land, construction, permitting and large scale social cooperation are all necessary while "road" vehicles works under anarchy, and the former have render projects as impossible in much of the world. The latter "technology" is superior and the former appears to attract precisely due to its dimension of moral superiority signaling, because it takes more cooperation and ideologically cooperation solves everything.

One should remember that a technology that just works is superior to one that requires society to jump over a dozen hoops to function. The moralists that equates constraints with goodness is weird.

The most damning part about rails is that the potential for practical improvement is out while other vehicle concepts are improving, at a point where rail is often uncompetitive to begin with. Regardless of who or how of a robot bus, costs are down and density is up.

For high speed rail systems, even places with existing systems is under intense competition for aircraft, and that is with cost and difficulty in building the track out of the way. New lines would basically never pay for themselves, and in the mean time electric aircraft is looking to greatly lower operating cost for short haul flight.
 
Last edited:
Its weird the train people would bring up high speed rail.
You mentioned commuter rail, that's the local commuter service. Starts a couple of stations up the line from me, in London in an hour.
 
The car (sorry the automobile) is inextricable from life in all but the centres of cities in the USA.
No Western European can grasp the size of the US and distances involved.
That said I am not sure Musk's toys offer much improvement over fuel efficient road vehicles. Nearly all adult Americans drive.
 
The car (sorry the automobile) is inextricable from life in all but the centres of cities in the USA.
No Western European can grasp the size of the US and distances involved.
That said I am not sure Musk's toys offer much improvement over fuel efficient road vehicles. Nearly all adult Americans drive.
'Tis said that Europeans think a hundred miles is a long way and Americans think a hundred years is a long time.
 
Most taxis are 3-4 adults, given they give up a seat to the driver. They do mostly operate carrying one or two people, but if you need one for taking the family somewhere, two seats isn't going to cut it.

The roadster styling is doubly problematical because it cuts down on available luggage space.
Hi,

All other things aside, while I guess I can understand the whole, most rides don't need more than two seats, like you and others I'm still left thinking that an extra seat or two would be very useful in enough situations to make it worth while.

For me, in the past I have used taxis for;
  1. Going to or from an airport or train station
  2. When going into the city when I would prefer not to take my own car (such as for a business meeting, or when I might be out with freinds and have a drink etc)
  3. When on travel to a city where I decided not to rent a car
  4. Going to/from the shop when my car is being repaired
  5. Going to/from the hospital or doctor's office
When using a taxi to travel either to or from the airport, train station, or in city business travel I will usually let the driver put my bags in the trunk, but will typically keep my computer bag/breifcase on the seat next to me. If I am traveling alone that wouldn't be a big deal, but if traveling with even just one additional person it could potentially be an issue. And in addition, for young families with kids a two person car just doesn't seem practical in any situation.

In addition to that, the car presented didn't really look very "mobility impaired freindly". In particular the seats seemed kind of low slung (more like a small sports car than a modern sedan or SUV) and the layout didn't seem to be well suited for someone in a wheelchair, on crutches, feeling physically unwell, or even perhaps someone with an injured arm or leg.

I guess an argument could be made that in situations like those above a person/group of people would have to hail a different cab, better suited totheir needs, but that raises a number of questions to me.

Looking at a lot of the newer "non-autonous" cabs that have been developed over the past decade or so, in most of them (such as the Nissan NV-200 NY City Cabs, the Toyota JPN Taxis, and the TX4 London cabs) there seems to be an effort to develop a multi use, highly flexible platform rather than a vehicle that seems to be focused more on a niche two-seat capability, even if that niche can accomodate a fairly large fraction of potential commutes.

And finally, as a side note, it strikes me as a little weird that while one of the purported benefits of an autonomus cab is that the passengers can do other things, the layout of the Tesla Cybercab doesnt seem all that well configured to the potential. Specifcially, the seating arrangement appears to be near identical to that of a small two-seat coupe, with the passengers seated side by side, facing forward, just without a steering wheelor pedals, whereas if you really wanted to allow the passengers the ability to do other things I'm left wondering why the seats don't face each other tomake it easier to talk toeach other and/or why isn't there a small table or tray where you can put a laptop or tablet wo that you can do work while the car drives itself. :oops:
 
In addition to that, the car presented didn't really look very "mobility impaired freindly". In particular the seats seemed kind of low slung (more like a small sports car than a modern sedan or SUV) and the layout didn't seem to be well suited for someone in a wheelchair, on crutches, feeling physically unwell, or even perhaps someone with an injured arm or leg.
It's a definite issue. Almost all the local cabs are Toyota Priuses, with a backend not to different from Musk's cybercab. My folded wheelchair will just barely fit in the boot of one, in one particular position, and only that position. Height is also an issue for getting in and out with a disability, normal height cars are awkward, low slung ones are really awkward. Two seaters do tend to be slightly better for manoeuvring things like crutches or casted legs into the seat as the door opening is typically slightly longer front to back than in a four seater. While the driver usually puts my chair in the boot (I suspect more to protect their paintwork than anything), I could do it if need be, but a lot of wheelchair users couldn't, and that's going to be an issue with driverless taxis (including aerial taxis!).
 
Last edited:

Please donate to support the forum.

Back
Top Bottom