Self-driving cars can solve "a" problem, just not the one you choose to formulate. Sure, we could probably make things safer by more tightly regulating drivers. But is that desirable? Presumably it is for you, since you likely consider yourself in the group of excellent drivers.

But I wonder if causality actually has any data at all to support his claim. I believe most accidents are less about driving skill and more about experience and attention. And young drivers have less of each. I believe accident data bears this out, as does personal experience.

Human error is very difficult to minimize. Computer error can be made extremely low with proper design, it's really just a matter of how much money we're willing to throw at it.

I actually enjoy driving and have a very good track record and I think well above average "driving skill". Do I think a properly designed self-driving car could be made with a significantly lower accident risk than myself? Absolutely.

You're assuming that their liability risk will go up. I suggest exactly the opposite. Sure, there will be accidents, but they will be far fewer than with human drivers. The rate of accidents where the computer controlled car is at fault will likely be 100 times lower. Even if bias against computer controlled cars will make lawsuits in those situations much more likely, and payouts much higher, I wouldn't be surprised if an analysis shows that their overall liability should decrease substantially.

In what way are we a long, long way away? If you're talking about an affordable driver-less car, I'd agree. If you're talking about laws being passed that allow their mass adoption, probably. But the technology is there. They can basically do everything you've suggested (home to work, detours, deer, kids), in many cases much better than people can.

Well, that may be true. But Waze seems to give you more options. I've not used it long enough to judge whether the options are bad, so I'll hold judgement. But when I try to reroute it seems to give more varied options, and thus far they've been pretty reasonable.

As others have pointed out, there are things that waze does better than google, such as the real time updates of map errors and traffic reports. Google maps has to wait for the traffic to start backing up, and that can occasionally screw you. Waze seems to sometimes be able to warn you before the cars really stack up.

If google could integrate some of these ideas in improving the speed of real time updates, it could drastically improve the usefulness.

Also, the rerouting functionality of google maps is pretty limited. There are very few routes it will tell you to choose. You can often figure others out by looking at the maps manually, but this is hard if you are by yourself.

If they can integrate some of the better rerouting and real time update tech from Waze, I could see it being a significant improvement to the overall experience. Would certainly keep them well ahead of the race. Sure, they're already drubbing everyone else, but in this business you need to kick them when they're down.

Whoooosh.

Actually I disagree. All you have to do is convince people there is a good chance it contains something of interest. That it is encrypted might entice people to download it in the hopes of discovering the key or decrypting it. I'll take your bet. I'll bet you a 2 TB hard drive.

I'm also willing to bet someone has blackmailed authorities into letting them go. I'm also willing to bet that said authorities did not announce that they were letting the accused go because he had some really juicy dirt on them.

Seriously? That POS physical button is the one of the few things I hate about my wife's S3. That and some of the Samsung customizations that clunk up the interface make me marginally prefer my Nexus despite the vastly superior screen and speed of the S3.

There is very good evidence for herd immunity for things like polio and measles. These vaccines are highly effective, and there has been very good epidemiologic evidence that 1) the vaccine is very effective 2) the disease is largely controlled when the majority of the population gets the vaccine, and 3) the disease beings to increase in frequency when vaccination rates wane.

The evidence is much less compelling for the flu vaccine. It is complicated to study because its effectiveness varies drastically from year to year, based on whether they guess right about which strains to immunize against. This year and last appear to have been a bad guess. Some years are better, but there are respected epidemiologists that argue that the evidence overall on the effectiveness of the flu vaccine in preventing hospitalizations and death is pretty weak.

A hospital certainly has the right to make policies they believe are in the interest of their patients and fire people who don't follow those policies. But to suggest that we should all be getting the flu shot because it doesn't do any harm is stupid. It is a pretty substantial cost to society and its use and effectiveness, like all immunizations and medical treatments, should be evaluated critically.

I'm not an Apple fan but I don't agree that Apple only targets novice users. Maybe you can make that case for iOS devices, but not OSX computers. There are plenty of advanced and technical users. I know plenty of engineers and techies who prefer Macs. I prefer Windows because that is what I grew up with and many of my computational modeling programs only work on Linux and Windows. But to suggest that only novices use Macs is silly.

It is not that necessarily that the weaker ones are killed off. But there has to be some advantage to breeding or survival for a mutation to increase in frequency in a population. Otherwise it just gets diluted out. It is possible that if a mutation isn't a disadvantage it could stick around as a lineage and then later thrive when some outside pressure gets put on the species such that it is now an advantage. So it may not need to happen right way, but it does at some point need to be an advantage for the whole species to evolve.

[I missed this and it is now old, but I'll respond anyway.] Actually, I completely got your point and tried to explain that you are confusing a mutation with cancer. Two points.
1. A single/few cell mutation that is destroyed is not cancer. By definition, cancer is an uncontrolled, invasive growth. If it is controlled right away, it isn't cancer.
2. I am saying that the technique would not detect these single cell mutations because the amount of DNA they would release into the blood would be minimal and the likelihood to detect in a random blood sample infinitesimally small. That is not a concern. The real concern is whether it can detect cancer early enough. Early cancer does not release as much DNA into the bloodstream as later, more invasive cancers. One would ideally like to detect it before it has spread all over the place, and it isn't necessarily clear that this technique will do that.

I really think you guys are worried about the wrong end of this. It is highly unlikely that this test is going to be too sensitive any time soon...quite the opposite, the key will be making it sensitive enough to be useful. One or a few cells aren't going to make enough DNA that you would have any reasonable chance of picking it up in a random blood sample. There would have to be some critical mass there already, and who knows, but I would guess that the amount of DNA released into the blood by even an in situ is going to be too small to detect.

It is true that cells mutate fairly frequently, but most of these are not "cancer". Cancer implies that it grows invasively. These sequencing tests would be looking for certain genes known to be linked with cancer. Perhaps over time they will develop heuristics that will allow for detection of mutations not previously characterized but initially it would probably be limited to cancer genes already understood. But my initial concern is whether early cancers dump enough genetic material into the blood for this to be useful for early detection. I'll bet that it is only after it becomes invasive that it releases enough DNA to detect.

Oh come on.
1) as pointed out above, this is less than a drop in the bucket. I would not call that a dent.
2) It is completely unclear if this would generate any net energy. A case can be made that many of these more inefficient biofuel processes consume more energy than they produce. How does that help.
3) Most importantly, things like this distract from the ONE thing that has a real chance at reducing our dependance on oil, which is nuclear. Solar and wind might help a little, and maybe biofuels can help with energy storage, but what is described here is not a significant part of any real solution.

You can talk about little steps here and there, but it is magical thinking. If we want to get serious about reducing gas usage (I'm not getting into whether this is the right thing, that's a whole separate topic), then nuclear has to be a huge part of the solution.

Actually no. Rate does not necessarily imply a percentage or ratio, it can be with respect to any other measure or unit. A rate can be per unit of time, which would be high. If the GP was using rate as per unit time, which is valid, then the statement is correct.