This already happens, in a human-driver world. We choose to continue driving our cars, even though we know there is a small but non-zero chance that on my trip to the grocery store, I'll kill somebody in the course of a car accident. If that were to happen, there might be a lawsuit involving my insurance company, and a dollar amount for the life of that person would be established. Factors that go into it would include the amount of negligence on my part, balanced against the amount of negligence on the part of the person who was killed. Is that life really only worth the settlement amount? No, you can't put a price on a life. But risk is a part of life, and we accept risks in everything we do. Even the risk of death. Self-driving technology doesn't change any of this, except for maybe who specifically ends up paying.

That's true. And the statistics tell us *exactly how commonly* these edge cases cause deaths. It doesn't matter how common the edge cases are, the only thing that matters is the number of deaths. Those robotaxis already have to encounter these edge cases every day, and yet the death toll is extremely low.

It's true. Why would we consider the statistics of bots in environments in which they don't operate? This is the real world we're talking about, not an extrapolation of what would happen if bots tried to operate in areas where they are not capable of operating.

Let's suppose bots take over 50% of driving in cities and highways, and just 5% of driving in rural areas. The lowered death rate brought about by the 50% of driving in cities, will greatly overwhelm the increase in deaths caused by the 5% of driving in rural areas. That's a win, even if bots are less safe than humans in rural areas.

You keep bringing up edge cases. It's true, there are situations that self-driving vehicles don't handle as well as humans. Construction zones are one of those.

But the statistics account for those edge cases. NTSB says that, over a one-year period, there were 6 crashes of self-driving cars that resulted in death. https://www.nstlaw.com/guides/.... SIX. That's compared to 41,000 caused by human drivers.

Overall, the chance of a self-driving car causing a death is lower than the chance of a human-driven car causing a death, per mile driven. That is true regardless of construction zones, off-road, or off-map driving.

Microsoft's security issues are a red herring, they have nothing to do with self-driving cars.

I disagree. AOL was, for many people in many places, the best internet service that was available to them, at any price. And for those who had choices, it was a reasonably-priced option. Even in recent years, AOL has served (mostly older) people who don't know how to jump through technical hoops to get "better" internet service, and it worked for them.

AOL's problem was that it failed to keep up with the times, wishing for the days when they *were* the portal to the internet.

Your nuance, changing "off road" to "unmapped" changes nothing. Those places are not where many human drivers drive, nor are they were many people are killed in any kind of accident.

Safety is all about statistical probabilities. If a new technology saves 1,000 lives and kills 10 people, it's very worthwhile. Take air bags. In 2024, air bags killed 17 people in the US. But they also save thousands of lives each year. That makes air bags a win.

If self-driving cars kill 10 people a year, but save 1,000 lives, then that's just as much a win for safety.

You are taking edge cases to prove the rule, and that's not valid. Just because humans are better off road, doesn't make them better overall. The vast majority of road trips are not off road, and the vast majority of deaths are not due to accidents that occur off road.

Your claim that 60% of humans are safer drivers than self-driving technology, is not backed by facts, it's just your own opinion. I provided a source, and can provide many more, they are not hard to find. Where does your data come from, exactly? Why is your number pulled out of thin air, to be trusted?

So because I support a technology that will *reduce* human fatalities from their current levels, I'm pathological? How do you figure?

Today, 40,000 humans are killed by human drivers in the US, each year. If robotaxis reduce that number to 30,000, how is that not a good thing?

So you're saying you like the higher rate of deaths caused by humans, and you wouldn't be in favor of robotaxis helping to lower the number of deaths?

Liability and acceptance are two different things.

We already *accept* that human drivers will cause a certain number of fatal accidents every year. We accept it in that we don't ban driving of cars, despite about 40,000 people being killed every year. That acceptance doesn't remove liability, for human drivers who are found to be at fault. Many are required to pay significant amounts of money, and many are prosecuted and charged with crimes, in connection with those deaths.

Acceptance of self-driving cars does not remove liability. The details of that liability are still being worked out through the passing of laws and the resolution of court cases.

First, the possibility that a few human drivers are safer than robotaxis, isn't really relevant when it comes to the overall safety of the technology.

Second, I don't agree with your premise on a couple of grounds.

Robots can indeed predict. Prediction engines are a key part of the architecture of self-driving technology. Further, self-driving technology incorporates sensory input that humans don't have access to, such as LIDAR, radar, sonar, and 360-degree cameras. These superhuman sensory inputs give self-driving cars an edge over even the best human drivers.

Also, unlike humans, robots gradually improve over time. If one human driver gets better at driving, other new human drivers are starting from scratch, learning how to drive for the first time. Robots, on the other hand, incorporate all of the safety improvements gained from all the "experience" of other robots, from the moment they start driving.

The numbers don't back this up. https://www.nature.com/article...

While your statements are true, this sensationalism won't kill robotaxis, just as EV fires make news, but haven't stopped EVs from gaining market share.

No, accepting that deaths will happen, is not the same as removing liability.

Today we accept that human drivers will cause a certain number of traffic fatalities every year. We accept it in the sense that we don't ban the driving of cars, not in the sense that we remove liability.

The acceptance of fatalities in the context of robotaxis means that we will not ban robotaxis, not that we won't hold the manufacturers liable.

Your statement, and my statement, can both be (and likely are both) true at the same time.

Those kids are already in danger from human drivers. If robotaxis reduce that danger, that is something to celebrate.

First, it does not follow that because Harvard caters to intelligent people, that the IQ distribution will not be Gaussian. The normal curve might not always start and end at the same point, given a certain subset of the population. But if the population is large enough, the normal curve will still apply.

Second, IQ does not correlate with all of the factors that make a student successful. It does not correlate with level of motivation to apply oneself, nor does it correlate with skills in every area. Just because you have a high IQ doesn't mean that you are good at everything.

Of course, the liability issues will have to be worked out through court cases and appropriate laws. Automated vehicles besides robotaxis aren't new, there are automated bus and train systems everywhere. Liability is a thing for them too.

Maybe you should try a restaurant that doesn't have bad or loud music, or babies screaming. Do a little Googling, there are plenty of good options. And even better, go with a friend. That's the real point of eating out...it's not about ingesting calories, it's a social experience. Try it, it's actually really good for you to go out and be with people, and not just online!