Put in that light, self-driving cars feel a lot like enshittification. The motivation isn’t the same but the effect is - our cars are no longer actually under our control. We just borrow control until someone out on the network decides it’s in their own interest to end the loan.
And meanwhile the true owners take every bit of data they can for their own profit.
This is where I land, except that I extremely desire a self-driving car. I would happily pay a substantial sum for that feature.
But I will never buy a car that I don't truly own, which means that I won't get a driverless car if the self-driving function is not fully contained within the car itself.
Only speaking for myself but I will never buy anything that is drive-by-wire which is a prerequisite for self driving cars. i.e. no physical linkage to/from throttle, brakes, steering. Too many hacking and take-over incidents on live highways already and I've spent the last few decades playing whack-a-mole on servers. Vehicles, ships and aircraft are the wrong place to fight this battle.
Drive by wire doesn't require networking. The F-16 is fly by wire and first flew in 1974. Certain cars have had partial drive by wire controls since the late 80s.
That's not what I said. Drive by wire is required for self driving. Networking is not required but just happens to be a thing on all of these cars. That is how they get their updates for self driving. Are you aware of a level 4+ self driving car that is not connected to the internet?
I suspect that even if there were physical linkage that would not be sufficient to give override control to the operator. The motors will be stronger than some operators and would require a lever that could forcefully disengage the drive-by-wire. Even that may not be fast enough as it just takes one quick yank of the steering wheel to send someone into an oncoming big-rig which the attacker will see on the camera which I predict will lead to an underground of on-demand market of snuff porn as there are cameras facing the occupants in some of these cars.
I’m pretty indifferent to self driving cars, in fact I still want a 5-speed for my next car, though I know someday I will go electric.
But I would really like my car to be self-PARKING. Why is that not a feature yet? Sure it can’t handle ALL scenarios, but can we get AI to handle mall parking lots and parking garages? Just drop me at the curb and put yourself away.
Several cars exist can self-park when you drive to the spot, but they are ridiculously slow at it. Add the basic self-driving functionality (much of which has existed for 5+ years even in mid-level cars like the Accord), and what you're describing is easily doable, though would cause massive backups in those garages I assume.
1. Like the article mentioned, bad media coverage. Trust that these systems will be truly self driving at some point seems to be dwindling. Enhancements may come, but I think people doubt that they'll be able to take a nap in their car anytime soon.
2. Sentiment for cars seems to be decreasing. A lot of people are realizing that cars are dangerous and advocating for more public transit, walking, and biking as safer and more affordable alternatives.
I used to think that self driving cars would solve safety, traffic, and give us more time back in our lives. Now I tend to think they add minor convenience for a good amount of risk for a high cost.
Risk to payoff is just too high, and the biggest problems drivers face are solved in many cars with a basic suite of assistive technologies (lane keep, dynamic cruise, etc). I usually drive with cruise control set to a practical speed and take over in basic moments (accelerating at green light, passing on freeway, exiting, etc)
Why do you feel That Tesla FSD is "buggy shit"? No, it's not perfect, but it is mindblowingly amazing. At the current state I'm quite happy with what I paid for it. I've driven over a million miles in the last 35+ years, it's quite nice to have my car drive me now.
I don't know, watching cars teleporting around doesn't I still me with confidence. That's just the most surface level problem I saw during my brief experience.
Wasn't it one of the CEO's of UBER in recent years that said they would never really be profitable until they had self-driving cars?
The end goal has got to be fully autonomous.
That could add up to a lot of financial pressure in that direction regardless of overwhelming public indifference or reluctance, as long as these countervailing forces combine to have a neutral effect.
I would imagine however big the amount of support there is for UBER (and things like that) to be successful, most strongly from the few biggest stakeholders, would then be pushing for full self-driving to no end, whatever it takes.
Anyone else who has to share the road with them, not so much.
17 comments
[ 0.18 ms ] story [ 54.5 ms ] threadPersonally, I'm neutral on self-driving cars. I mean, I don't find the prospect of owning one either compelling or repelling.
But self-driving cars seem to come hand-in-hand with being connected to a server somewhere. I find that repellent and would never buy such a car.
But I will never buy a car that I don't truly own, which means that I won't get a driverless car if the self-driving function is not fully contained within the car itself.
That's not what I said. Drive by wire is required for self driving. Networking is not required but just happens to be a thing on all of these cars. That is how they get their updates for self driving. Are you aware of a level 4+ self driving car that is not connected to the internet?
I suspect that even if there were physical linkage that would not be sufficient to give override control to the operator. The motors will be stronger than some operators and would require a lever that could forcefully disengage the drive-by-wire. Even that may not be fast enough as it just takes one quick yank of the steering wheel to send someone into an oncoming big-rig which the attacker will see on the camera which I predict will lead to an underground of on-demand market of snuff porn as there are cameras facing the occupants in some of these cars.
But I would really like my car to be self-PARKING. Why is that not a feature yet? Sure it can’t handle ALL scenarios, but can we get AI to handle mall parking lots and parking garages? Just drop me at the curb and put yourself away.
How far away is THAT!
1. Like the article mentioned, bad media coverage. Trust that these systems will be truly self driving at some point seems to be dwindling. Enhancements may come, but I think people doubt that they'll be able to take a nap in their car anytime soon.
2. Sentiment for cars seems to be decreasing. A lot of people are realizing that cars are dangerous and advocating for more public transit, walking, and biking as safer and more affordable alternatives.
I used to think that self driving cars would solve safety, traffic, and give us more time back in our lives. Now I tend to think they add minor convenience for a good amount of risk for a high cost.
The end goal has got to be fully autonomous.
That could add up to a lot of financial pressure in that direction regardless of overwhelming public indifference or reluctance, as long as these countervailing forces combine to have a neutral effect.
I would imagine however big the amount of support there is for UBER (and things like that) to be successful, most strongly from the few biggest stakeholders, would then be pushing for full self-driving to no end, whatever it takes.
Anyone else who has to share the road with them, not so much.