Yeah, I looked it on Fox Busness Network, and yeah it wasn't the car that did the incident, probably a misprogramming of the "AI" (Artificial Intelligence) that didn't recognised a human walking on a crosswalk in a intersection.

This just shows you that how bad self-driving cars are, and yet can cause a hazard. Correct me if i'm wrong on this.  No offence, ok?

The car did, what was programmed to (driving straight, when it sees no obstacles on its sensor systems). But it was the driver and driver only, who spectacularly failed to do his duties: Observing the situation, so he failed to spot the person and take over the control and so failed to evade the collision.
The problem is the same as what is behind vast majority of accidents: Driver doing whatever else, but not driving (with cat 3 it means supervising and be ready to take over the control). This is about the phones, navigation and many other distractions and generally the lack of drivers discipline to really focus on the driving and not letting the self being distracted.

Concerning cruise control, I heard a story of a guy who had just bought a top of the line Winnebago with cruise control. He was very pleased with it and set off on a road trip, wanting to make himself a cup of coffee he put the cruise control on and went into the galley to make his coffee. That year he won the Darwin Award because he crashed off the road and died. Cruise Control is not the same as auto pilot. 

For auto pilot, that would be for a airplane.

A nice thick wire connecting the accelerator to the carburettor is the only drive by wire I want. 

Carburetor! Now that is a blast from the past. My first car had a carburetor.

For auto pilot, that would be for a airplane.

Exactly, but apparently this guy didn't know the difference. 

Exactly, but apparently this guy didn't know the difference. 

Well, there is way less difference than you think. What is called "autopilot" in airplanes is nothing more than a set of assistants not much different in capabilities than the cruise control in the car. The main difference is, in cars 90% of the drivers set the automated feature and blindly rely on it, but in airplanes 99% of the pilots set the automated feature, but then keep constantly monitoring what is does. Exactly the same as the drivers should do. Going to galley to fix coffee is not monitoring what the machine does.

And by the way the Uber case is even different than these: There the lady stepped out from the shadow directly in front of the car. So from what/whoever drove that car, the collision was just plain unavoidable. Some even speculate the "autopilot" did manage to at least slow down a bit (so increased the chance of survival) compare what a humen driver would be able to do in that situation, but unfortunately it was not enough...

It was the Tesla case few years ago, where it was really 100% humans driver (the victims) responsibility...

Carburetor! Now that is a blast from the past. My first car had a carburetor.

My last bike had four 

Short answer: nope.
Long answer: besides the problems of systems failing and/or being hacked, the driver cannot be depended upon to remain attentive when just observing.

... the driver cannot be depended upon to remain attentive when just observing.

Then he/she should not use the automated feature at all.

In aero and marine industry the "set the auto and then monitor it" works for long years, I see no reason why the same can not work in automotive as well.
There are many tasks the machine can do well and in many cases way better than humans (shortly said as all needed to maintain the selected track in whatever conditions, evade known obstacles; practically all what needs fast and coordinated response), but there are many areas where the machine severely lacks and will lack for still very long time and where humans are way superior (selecting the correct track on the road, evaluating what from the surroundings is a thread and how big).

With commercial ships and aeroplanes all construction and namely the training goes in the direction of utilizing advantages of both and so make them cooperate. Even the "pilot less" aircrafts/boats (today mainly the military drones, but many companies want to bring that to the cargo transport) are by far not "pilot less" at all, the "pilot" is just physically not present in the vehicle and sits in some distant office, but he still does the same job: Monitors the machine, the surroundings and commands the computer for the exact route and actions (e.g. navigaes around/through bad weather for the mission,...).

Who needs to be kicked to their a$$es are the marketing bu**$$tters, who are brainwashing the public "in a selfdriving car you do not have to drive it" way. Yes, it sells those features, but it is this nonsense, what is responsible for most of the accidents involving driving automation.
What I see as very bad approach is to hide from the customers the exact ways, how the automated systems make their decisions (e.g. a thing as primitive as when exactly the Start/Stop feature shuts down the engine and when not; manual really lists "for the driver it is unpredictable"...)

Then he/she should not use the automated feature at all.

In aero and marine industry the "set the auto and then monitor it" works for long years, I see no reason why the same can not work in automotive as well.
There are many tasks the machine can do well and in many cases way better than humans (shortly said as all needed to maintain the selected track in whatever conditions, evade known obstacles; practically all what needs fast and coordinated response), but there are many areas where the machine severely lacks and will lack for still very long time and where humans are way superior (selecting the correct track on the road, evaluating what from the surroundings is a thread and how big).

With commercial ships and aeroplanes all construction and namely the training goes in the direction of utilizing advantages of both and so make them cooperate. Even the "pilot less" aircrafts/boats (today mainly the military drones, but many companies want to bring that to the cargo transport) are by far not "pilot less" at all, the "pilot" is just physically not present in the vehicle and sits in some distant office, but he still does the same job: Monitors the machine, the surroundings and commands the computer for the exact route and actions (e.g. navigaes around/through bad weather for the mission,...).

Who needs to be kicked to their a$$es are the marketing bu**$$tters, who are brainwashing the public "in a selfdriving car you do not have to drive it" way. Yes, it sells those features, but it is this nonsense, what is responsible for most of the accidents involving driving automation.
What I see as very bad approach is to hide from the customers the exact ways, how the automated systems make their decisions (e.g. a thing as primitive as when exactly the Start/Stop feature shuts down the engine and when not; manual really lists "for the driver it is unpredictable"...)[/i/]

I agree with you 100% on this, automated processes that assist the driver are the best option in my opinion.

I agree with you 100% on this, automated processes that assist the driver are the best option in my opinion.

Yes. ASSIST, not take over. Even in commercial aeroplanes, the autopilot is there to assist the pilots, not take over completely. Some people seem to think that they will be able to just get in a "self driving" car and tell it they want to go somewhere like they would in a taxi. That's dangerous; there is no way that an automatic system will be able, under normal road conditions, to take over control completely.

If those things ever appear on the roads around here, I'm handing in my driving licence, no way do I want to be on the same piece of road as them, thanks all the same 

I’m good with cruise and abs, though I don’t need either. Heck, I don’t even need AC and I live near ATL! Even now I don’t use it that much even though I do have it. I don’t like systems automatically intervening on my control of the car that much. They need to have an off switch.

Heck, I don’t even need AC and I live near ATL! Even now I don’t use it that much even though I do have it.

And how do you keep the "climate" in the car bearable?
I ask, because there is rather wide spread myth "The AC is increasing fuel consumption".
Yes, running the AC needs some energy, guesses are 0.2l/100km or so.
But driving at high temperatures severely influences the peoples reactions. That means you will be responding rather late, so instead of assessing the situation well in advance and slowing down earlier by just releasing the "gas", you keep feeding the engine till last moment and then burn the excess energy on brakes. Costs more than 1l/100km on overall fuel consumption. Not yet speaking about safety...
Or using open windows? Well, the open windows use to upset the aerodynamics, leading to about 0.5l/100km extra.
So the 0.2l/100km is extra energy, but that is still less than both other "options"...

I don’t try to make the climate in the car to suit my needs. It’s my goal to be adapted to the outside climate to be prepared for whatever comes my way. Driving in the car is a convenient time to climatize. I live and work in a wide variety of conditions and I like to be relatively comfortable wherever I go.

As for fuel economy goes in my experience driving habits and traffic have more impact on fuel economy than whether or not the ac is on. Lastly reaction times are dependent on many of factors including the ones you mentioned. Since I am most comfortable in high heat environments I believe this to be a negligible impact at best. At any rate I believe that competence is more important than safety as you can never create a 100% safe environment that is healthy for people, but you can learn to be competent to deal with whatever comes your way. To that end I am constantly honing my driving skills to be prepared for the unexpected an addition to keeping my vehicles in good shape to deal with whatever comes my way. It has served me well so far.

The further driving automation is implemented, the less competent people will become at driving. People driving aren’t paid pilots and boat captains that require massive amounts of theory, training, and practice seat time to operate cars. They’re regular joes and janes just wanting to get to work, etc. it’s not a good comparison. There’s not really to my knowledge a direct comparison of a privately owned and operated domestic transportation means that has gone to full automation for us to look at yet.

Maybe this was mentioned previously, but I want to mention "infrastructure". I work in a DOT, and the current mentality of many DOT's I know of are in a "deferred maintenance" and "how cheaply can it be done?" mentality.

Driverless cars require a few things from the DOT:
-clear lane markings
-consistency along path, including lane straightness, curb heights and standardized and consistent spacing to fix objects/barriers.
-good (at least) road surfaces with consistent traction profiles and no potholes
-clear and advanced communication of planned or emergency roadwork
-clear and advanced communication of path alterations
-amongst other challenges

See, DOT's in USA currently are unwilling to provide these requirements because the people in charge are getting very lax in doing their jobs.

Driverless cars use cameras to keep in lane, driverless cars rely on inboard data to predict a path that can't be altered unless the car is updated, GPS can only get a car going in "generally" the right area but the car will rely on sensors to fine-tune it's position.
Road surfaces with varying coefficient of friction (including rain and ice) will be a huge obstacle for autonomous vehicles.
Bad weather another obstacle.

Parking...hmmmm, I had a loaner car that can park itself. Cool tech, and been out for a while now. So, tech may have a good solution here, EXCEPT, how does a driverless car know WHERE to park, and when (TOD restrictions)? Autonomous cars want to be without steering wheel or pedals, so how can a passenger/operator recover a car that enters a non-specified error state that it can't get itself out of?

Insurance. Here I truly believe defects in sensors, buggy software, or missing abilities in the vehicle (ability to predict someone running out from between cars or when a semi pulls sideways in front of car, a Tesla famously killed its passenger in this scenario) will mean some amount of liability will undoubtedly fall back on the vehicle manufacturer, especially in bigger cases. If it is a Ford autonomous vehicle, and the vehicle makes a mistake, doesn't that mean Ford is responsible (especially in a fully autonomous vehicle)??