There's absolutely zero reason to believe anything the NSA says about how Snowden got the documents, or indeed, about anything. They believe they are entitled to lie to congress, so the public isn't even a question.

"If the villagers didn't riot and just left quietly instead then you would still be left wanting better comments."

No, actually. There's more than enough in line to replace your oh-so-high quality comments. Good riddance.

What sank digg was not UI changes, but handing over control totally to advertisers. UI changes are not something to act disruptive in every fricking thread over, even if they reduce features. Slashdot is based on antediluvian perl code and has struggled with it for ages, it's not surprising that some features are disabled - permanently or not - in yet another attempt to modernize it. As long as they keep the archives, and maybe one day even make them effectively searchable, they can do whatever they wish as far as I'm concerned.

And as far as I'm concerned, the beta protesters are just a bunch of bored attention seekers.

There's a reason why you can't just strip out FAT support, a reason those patents are so obscene. It's a de-facto standard, and you need it for compatibility with lots and lots of stuff.

The actual technical worth of the FAT filesystems is zero. They are dumb, slow, they fragment, and lacks essential features. You can have strictly superior systems for free. But due to network effects, it's very hard to get rid of as the lowest-common-denominator filesystem, that can be read on every Windows and OSX and dumb little flashcard-reading gadget.

But can a chat bot plus big data really produce anything beyond a creepy, awkward facsimile?

No, it cannot. Once you're dead, you're dead. Game over.

Sure it can! It can provide notoriety, media attention, and maybe even a position in the ethics faculty.

Seeking a false balance between the truth and the lies, is a common strategy when the lies have failed.

Yeah, it's getting warmer. And OK, it's getting warmer faster. And although it's less clear, it seems this is speeding up too. But if you look at the third derivative, it's not so clear, there might actually be evidence of a slowdown!

("retreat to the derivative", a common strategy in numbers-based arguments...)

The indigenous populations of the Americas (north and south) was somewhere well under 112 million prior 1492

That is such a hopeless argument. For one, you do not need a lot of humans to drive megafauna extinct. Not any more than you need a lot of snakes to drive flightless birds extinct on an isolated tropical island.

More importantly, this happened long, long before 1492! Some 13000 years before. The population in 1492 could have been a billion or it could have been one for that matter, but either way still wouldn't have proven or disproven the overkill hypothesis.

I never believed the hunted to extinction nonsense.

People, especially anthropologists who live with native people and/or devote themselves to try to understand their way of life on a deep level, are understandably reluctant to believe such disheartening things about their study subjects. Some of them argue passionately against it, but that is one of the times you need to look a little critically on their actual arguments - especially the things they can't explain.

If humans did not contribute decisively to killing off American megafauna, it was a fantastic coincidence of timing.

Yes, there's always questions about the accuracy of rankings/ratings in a system where people can choose who to play. But if only lower dans are willing to play the top bots, that speaks volumes in itself.

In general, since players are protective of their ratings and pick and choose who to play, while the bot makes no such considerations, I'd say the bot is probably stronger than its awarded rating indicates.

Well, loving is hard to measure, but it's true: If your single mom never (re)marries, or marries someone else than your dad, that makes statistically no difference for your life outcomes.

Adopted kids don't do too badly - but that's because prospective adoptive parents are heavily screened (and thus more likely to be loving and stable, sure!) not because biology is unimportant.

They might glance at your move, and they instantly know that the move was never played in a published game, but that GM So-and-so evaluated the move in his analysis of another game between Foo and Bar, where a different move was actually made.

Yeah, Carlsen is very impressive in this regard. But remember, while this might seem an insane feat of memory, it's made possible primarily because the moves (and the discussion of the moves) makes profound sense to the top level chess players. It's not like memorizing a random sequence.

Even a little bit of randomization shakes up their assumptions. I don't know for chess, but for Go (which is much harder for computers, and computers are still only at a high club level of play) there have been experiments with that. If you instead of starting with an empty board, start with, say 16 random moves, you get a position from which the computer will trounce even Go pros - regardless of color. The Go pro's understanding of the game is so tied up in situations that are likely to arise from sensible play (from at least one player), that they evaluate random positions far worse than computers.

It's well known that Carlsen isn't super-thrilled at these sort of exhibition games with unranked players, so Gates is just being nice by playing quickly. He'd lose anyway...

And if you inserted a cassette with Bubble Bobble, you only had to press play and wait about 15 minutes for it to load.

So I have to show my work, but you don't? A classic of the time- wasting troll.

Many Faces of Go is a decent program today, and back in its pre-Monte Carlo days it was also decent - as computer go programs went. Version 11 (the last pre-Monte Carlo) I have no trouble believing a non-club player beating. The later versions, well, they're certainly within range for an amateur to beat too (1-4 dan on KGS, depends a lot on the power of your computer) but then your dad is pretty good :)