I am less concerned about the potential for Chinese interference than about how American firms such as Palantir are embedding themselves in the structures of government and daily life. Through this they are collecting data on all of us which they will use to monitor and control us in the interests of their authoritarian owners and their political allies.
A safety fallback that can stop a vehicle is not a ‘kill switch’ in the ominous sense: it’s a fail-safe.
The real question I believe is, whether the system is transparent, redundant, and recoverable. And if we’re going to worry about that, we should apply the same standard to every robotaxi operator, not just the one the Chinese one that you deem suspicious.
Switching off the power supply of a heavy vehicle that is in motion isn't a "fail-safe". A fail-safe is a mechanism that, in the event of a system failure, leaves the device in a safe state. For instance, magnetic locks on building doors disengage when power is cut, which is likely to happen during a fire, to ensure doors that are held shut can be opened for escape, and fire doors that are held open close to provide containment.
A killswitch is something that intentionally prevents the operation of a device by some means of activation, with or without regard to what that device is currently doing, which can make it dangerous.
Whether or not Chinese providers are being singled out as a specific concern for potentially having these killswitches is a different matter
In a moving autonomous vehicle, a fail-safe does not have to mean preserving motion. It can very well mean a controlled stop, pulling over, or another fallback that leaves the vehicle in the least dangerous state.
The real question in my opinion is not whether the system can be stopped, but whether it is designed to handle outages and faults safely and predictably. That is a legitimate engineering issue.
What it is not, by itself, is proof of some sinister “kill switch” in the sensational sense.
The article describes a device that could remove power from the vehicle, that's a kill switch. The question of fault tolerance is one for all automated vehicles and isn't the subject of this piece
A power-removal mechanism is not automatically a scary kill switch.
In autonomous vehicles, it often is part of the safety architecture. The existence of a shutdown or disconnect mechanism is not evidence of a potentially dangerous flaw. In fact, I’d argue it is evidence that the system was designed with fault handling in mind.
Imho an autonomous vehicle, not having that capability would be the alarming scenario.
I am less concerned about the potential for Chinese interference than about how American firms such as Palantir are embedding themselves in the structures of government and daily life. Through this they are collecting data on all of us which they will use to monitor and control us in the interests of their authoritarian owners and their political allies.
A safety fallback that can stop a vehicle is not a ‘kill switch’ in the ominous sense: it’s a fail-safe.
The real question I believe is, whether the system is transparent, redundant, and recoverable. And if we’re going to worry about that, we should apply the same standard to every robotaxi operator, not just the one the Chinese one that you deem suspicious.
Switching off the power supply of a heavy vehicle that is in motion isn't a "fail-safe". A fail-safe is a mechanism that, in the event of a system failure, leaves the device in a safe state. For instance, magnetic locks on building doors disengage when power is cut, which is likely to happen during a fire, to ensure doors that are held shut can be opened for escape, and fire doors that are held open close to provide containment.
A killswitch is something that intentionally prevents the operation of a device by some means of activation, with or without regard to what that device is currently doing, which can make it dangerous.
Whether or not Chinese providers are being singled out as a specific concern for potentially having these killswitches is a different matter
In a moving autonomous vehicle, a fail-safe does not have to mean preserving motion. It can very well mean a controlled stop, pulling over, or another fallback that leaves the vehicle in the least dangerous state.
The real question in my opinion is not whether the system can be stopped, but whether it is designed to handle outages and faults safely and predictably. That is a legitimate engineering issue.
What it is not, by itself, is proof of some sinister “kill switch” in the sensational sense.
The article describes a device that could remove power from the vehicle, that's a kill switch. The question of fault tolerance is one for all automated vehicles and isn't the subject of this piece
A power-removal mechanism is not automatically a scary kill switch.
In autonomous vehicles, it often is part of the safety architecture. The existence of a shutdown or disconnect mechanism is not evidence of a potentially dangerous flaw. In fact, I’d argue it is evidence that the system was designed with fault handling in mind.
Imho an autonomous vehicle, not having that capability would be the alarming scenario.