So why the hell do we allow Google to release non-working beta software on the internet?

Because:
(1) Google's "non-working Beta software" often works better than software from other companies that purports to be ready for general release,
(2) For most software purposes on the internet, there isn't the kind of immediate public safety concern that justifies regulation of what vehicles are allowed on public roadways.

Why do we allow them to blatantly violate our privacy and sell our information to advertisers? It's time for the government(s) to step in and do something about them.

Insofar as thee have been actual credible accusations of privacy violations at Google, governments -- both in the EU and the US -- have stepped in.

If you have information on cases where that has not occurred, you should provide specifics, rather than vague handwringing.

Though, preferably, in an appropriate place -- even if you had a point, without some nexus beyond a connection to the same company, it would still be off-topic in a thread on Google's driverless car technology.

Google needs to be shutdown, or at least they need to be made to change their blatantly obvious and abusive business. And if they refuse, the CEO's and higher level people need to be put for jail for their violations.

As a pretty firm believer in the principal of legality as opposed to the rule of lynch mobs, I'd like to see some credible evidence that the "CEO's and higher level people" actually committed offenses for which jailing is the punishment prescribed by law before accepting that they ought to be put in jail.

My guess is because the license will be granted with the understanding that it's a research vehicle. Someone will likely want to be closely monitoring the output of the car's instruments, so this insures one guy can do that while the other focuses on the road.

If there wasn't this requirement, one guy could conceivably monitor the instruments and not pay attention to the road since the car is driving itself.

Why would he be particularly immortalized? For example if you're looking for the first human to be killed by robots, you don't have to wait for "I, robot" to become reality as that happened already back in 1979 [wired.com]. Doesn't mean that robots have went away, people are quite regularly maimed or killed for neglecting safety zones, getting caught in presses and grinders and such. My prediction is that the first person killed by a computer-controlled car will be a Darwin Award winner that would have been killed by a human driver too, had there been one. Don't get me wrong, a computer-controller car won't be better than the people who programmed it and it surely will have bugs, but that one can be refined and get better whereas today every year we let loose a new generation of unskilled teens on the road.

Perhaps the best analogy is healthcare, you know those life-and-death situations you'd think keep everyone on their toes constantly. Well, nurses and doctors are humans too and they make mistakes, not often but they do. Electronic systems that make sure people always get the right medication in the right dosage at the right time, that they don't get dangerous combinations or medicines they're allergic to has helped save lives. Start counting the times the system corrects the nurses versus the times the nurses corrects the system and you'll find out who is actually the more important part of the two.

And that's why I think computer driven cars will win out in the end, they will always stick to protocol. They'll obey all speed limits, keep distance to those in front, always change lanes cleanly, always signal, always yield, always drive defensively and eventually all the accidents that don't happen because a human was tired or angry or sloppy or fiddling with the radio or his phone or whatnot will outperform the "creative" thinking capability of humans. Our ability to make good split-second decisions in an emergency situation is overrated, not to mention the choices are rather limited to break, turn and possibly in a few situations give gas. Many people panic and actually make it worse than just slamming the brakes.

I expect these cars also will have the ability to record near-accidents which you can use for analysis, you don't actually have to have an accident. Here we just managed to perform an emergency brake for a pedestrian who suddenly walked out into the road, could we have done better? Was our response optimal given the data we had? I see a whole new level of preventive improvement possible here. There's no significant learning for me from having one incident every decade, but if you can collect thousands of situations from millions of drivers it can learn to handle the 0.01% situations that we never have any training for or guidelines for what to do.

My personal feeling is that insurance rates are going to drive the adoption of self driving cars. Once the insurance companies realize that they have a lower error rate than humans (never tired, drunk, distracted, etc) and that they can tell who was at fault in an accident (almost certainly the other guy) you'll see serious incentives to keep cars in auto-drive.

If there wasn't this requirement, one guy could conceivably monitor the instruments and not pay attention to the road since the car is driving itself.

Yet half the drivers on the roads in Las Vegas, Nevada are TALKING ON THEIR CELL PHONE instead of paying attention to the roads! Even after it was made illegal! And many of them can't even maintain a normal speed or sometimes even lane position! Nevermind being able to react to anything around them.

The penalty is a fine, but perhaps it should be having the phone inserted via one's rectum. (*)

Scratch that, too many weirdos here would like that.

(*)

"Hey that driver with the cell phone just crashed into an innocent person.. They should shove that phone up his ass"
"You mean rectum".
"Rectum? Damn near killed 'em"