Chip clock speeds have sort of hit a ceiling, so beyond-factory overclocking doesn't do as much good anymore. What are you going to do with 10% more clock speed nowadays? If playing games, you'd just upgrade your video card. If you were really serious about overclocking, you'd water cool. There's just not much reason to eek out a tiny bit of cpu performance with slightly better air cooling. The stock Intel heatsink and fan is quiet and performs well, and there's not much reason to spend more money.

Yes, you buy that. I will be in the 3D-printed car that launches the deer over to your car for a 50% cut of the meal.

If we had truly autonomous cars, we wouldn't need a car per person. One car can take you to your park-and-ride, your wife to work, your son to middle school, and your daughter to elementary school an hour later. Then, it can pick each person up and take them home. And just in case scheduling conflicts, you can team up with your brother and sister to form a 3-car system. Team up with more people, and you can start carpooling and sending the nearest available car to whoever needs it like a taxi service. Get a city involved, and you'll have the more adaptive and cheap bus system in the world, that picks you up on your doorstep and transfers you from car to bus with perfect timing. Routes and transfer points will change dynamically to route traffic most efficiently. Bus-only lanes and traffic light control will ensure calculations are accurate for the majority of the route. Even if you drove like a maniac, you'd have trouble beating an autonomous system that synchs all the traffic lights to its benefit, drove speed limit on the bus-only lane, and does a perfect transfer to car to take you from doorstep to doorstep. Or maybe it wouldn't be that hard because there will be so few cars on the road that owning a car would be like having your own private Jet.

That's what the uranium is for.

I find the monthly bandwidth limits a bigger problem. Most months we don't use much, but one or two times a year, we go over.

He didn't get fired. He was offered a lot of money to be CEO of Zynga, then the Zynga stock immediately went up. I wonder how long before they realize what a huge mistake they've made.

The author thinks MS should make a low-cost tablet, but I think that's just further suicide. The low-end is glutted with tablets, lead by sellers who are making next to nothing on hardware to push their marketplaces and followed by their OEMs. Microsoft is not going to be able to undercut them (would be lucky to just match, given the quantity discount disadvantage), and then they will be in the same fiasco as now. What's been proven, and what people have said since day one - the Surface Pro will sell, and RT will not. You don't enter markets where you don't have a differentiator or a price advantage. A cheaper Surface Pro on Intel's future mobile x86 chips might be a good route to take (even Samsung is buying into them). They won't make it big that way, but there's nothing wrong with just making lots of money.

... bring a doggie treat

The comparison to Amazon/Yelp is misleading. They are confusing the rating of the comments (e.g. the "most helpful" positive/negative on Amazon) to the product rating. Upvoting a comment on Amazon does not change the product rating, and it doesn't on Yelp either. A comment does not become more trustworthy with upvotes - it just gets possibly seen more. If the researchers wanted to study Amazon ratings, they should have written fake positive reviews on Amazon products. Regarding the trustworthiness of comments, that's very hard to study - are you going to track down users and give them a lie detector test? It's almost like they realized their study was lame and tried to extrapolate it to a more exciting conclusion after-the-fact.