Assorted links


It's your imagination -- Mitt was going after Obama, and a bit of media, NOT Catholics.
"Obama was embraced by Catholics, Romney had dinner with rich folk".

The letter O, the number 16 trillion.

"So little time, so much to redistribute"

Thanks for a great link. What lines did you think were pokes?
Obama is more vulnerable to the redistribution poke, but many Catholics seem pretty happy at gov't based redistribution, too (unfortunately).

Holy crap, I haven't been following chess for a while now. I thought Rybka was God. To see it trounced like that...

It's like that thing in the anime series DragonBallZ. After four seasons, the main character finally becomes Super Saiyan, the strongest warrior in the universe. Then a bunch of the supporting characters go Super Saiyan as well. And then the show starts introducing Super Saiyan level 2, 3, 4 ...

I shudder to think how powerful Houdini 4 will be.

The game was magnificent, surely an instant classic. Worthy of Morphy. One question, though - I don't understand 19.f3. That seems to be where things begin to go wrong for White. Surely it would have been better to take the material advantage and leave that square for the knight. He should have played Nc2-d4 Kc2, protecting king and occupying center. (The benefit of hindsight.) I refuse to believe that Houdini was winning from the first gambit. Haven't put it through an engine yet, though... (lol)

Seems to me that White should have put a lot more effort into defending b2-c3, instead of abandoning it, despite the strong pressure, given that there was really no option but for the king to hide behind there.

It just occurred to me that it plays exactly as described by "the Formula" in the movie Revolver


It's interesting that Houdini is both better and less drawish. My impression was that that was seen as a trade off, although I know a lot more about computer chess programming than chess itself, so that might be an incorrect impression.

It's also nice to see that progress in computer chess persists at a fiercely high pace despite so much of the game community's effort being put into more complicated games. The fact that chess didn't just die 10 years ago when the machines became unbeatable is a positive thing, I think.

Have we yet reached the stage that was predicted (I think it was Seirawan or Nunn) where grandmasters are regularly thrashed by their cellphones?

Heh, I don't know. I don't think it's common for grandmasters to have serious matches with machines anymore. The best phone chess apps claim world-class ELO ratings, and they are probably right to.

My understanding is that the best humans in the world need better than knight odds to compete at normal time controls with PCs. I think humans+computers still beat straight computers in correspondence chess.

> I think humans+computers still beat straight computers in correspondence chess.

You know, I said this, but looking around on the internet I don't see any evidence of this actually being tested at a high level in chess. It just seems like another one of those human-biased myths games players propagate. I could believe it, not least because things like Houdini are not optimized for month-long runs, but I could also believe that it's just pride talking.

Kasparov would be very upset to hear you talking like this...

> Kasparov would be very upset to hear you talking like this…

There's a great paper by Jonathan Schaeffer called "The Games Computers (and People) Play" which goes through the history of computer game playing, though it's a little dated. It's chock full of tales of hubris. Chess is too positional. No computer will ever defeat humans in bridge! Crossword puzzles require human understanding. It always ends the same way. :) You should hear the Go people talk.

Is the future of Chess Algorithm-vs-Algorithm? Will humans slowly retire from active game playing? That point doesn't seem too far away....

I don't know that this has occurred. If anything, I think that the wide availability of great chess programs has made it easier to get great at the game. But someone more in tune with the chess community would have better insight than me.

If it wins more and draws less, it makes me think that Black might have a winning strategy. Perhaps moving first is a significant disadvantage, particularly in a game where you must move a piece and half of the pieces can't move backward. White is forced to open up before Black.

No, it's easy to waste a move. The game of chess is mostly likely an objective draw, otherwise it would be won for White.

Houdini is just better.

Empirically it looks like White has a slight advantage winning ~54% of non-drawn games. So it's hard to believe that moving first is a disadvantage.

But yes, that probably is the best computer chess game ever, even ignoring the players.

+1, and I would go a little further and note that as the commentator said, this was really more akin to a Morphy-like romantic match. Unbelievable as you see the match unfold trying to figure out what Black was going to do and then all of the sudden the hammer comes down. I think we know why Kasparov retired. Well played match.

Oh yeah, thanks for that. One of the most entertaining games I have ever watched.

I'm no great shakes as a chess player and I haven't followed progress in machine chess for many years. As I followed through this match, I tried to pick the best move. I only picked two moves in the whole match. It was so original and unexpected that I spent a lot of time thinking "Why did it do that?" It was fascinating. This is my first encounter with computer chess for a long time and it really hammered home why grandmasters can't win any more.

Just tried to make Romney tax plan work. If the applet is correct, it seems you can recoup the revenue, without breaking any promises, although you need a much smaller cap on deductions than 17000. And that's a static model, so growth could make it even easier. So I guess the WaPo should abanon its "mathematically impossible" stance.

Those were some great speeches. Who else though Romney was slightly funnier, though?

...even a cap at 17k, let alone much lower, will amount to a massive middle class tax increase


Lowering tax rates on the middle class is not the same as lowering taxes for the middle class.

Romney's plan is a giant transfer from the poor and middle classes to the upper class; from social programs to big business and the military industrial complex.

His "plan" won't reduce spending, and it certainly won't reduce the deficit. This is all just a redistribution scheme.

How is he going to give money to the rich?

I didn't use the word "give" because from a tax perspective the government is not in a position to do any giving, being a collector and all. Just as "lowering tax rates" doesn't equal "lowering taxes": words matter.

That said, his plan is to swing even more of the short-term - and especially long-term - financial responsibility of funding the government onto the backs of the lower and middle classes.

"Even more"? You realize the U.S. income tax is the most progressive in the world? 50% of the population pays no or negative income tax!

That said, your claim is horseshit like all the others about Romney's plan. We don't know what the plan is, we only know the general parameters. And he absolutely can achieve all the parameters he set without increasing taxes on the middle class (which is one of the parameters). So why pretend he's "really" going to violate that parameter? Claims about which ultimate reforms are more or less politically impossible are just confirmation bias. If you want reality, most likely none of it is going to happen anyway even if he gets elected.

People ragging on Romney, have you all read Feldstein and Rosen on this? Which bits do you disagree with? At any rate I think the "mathematically impossible" rhetoric has been successfully blown out of the water, for those who still care about objectivity. (And that includes the distributional neutrality.)

Brad Delong pretty much refuted Feldstein's analysis iirc. Doesn't change that Delong's a dick.

Tyler missed the real story:

The culture that is Japan.

Also, there is no great stagnation.

Subramanian seems to have a pretty narrow concept of "politics conducive to good economic institutions" (democratic/non-democratic). As an Indian, he should know better than to blithely assume that Indian politics are more benevolent than Chinese. (They are on the whole, perhaps, but not when it comes to maximizing output.)

re: 4

What a historical irony it is that the moon landing, the US' victory against the collectivists, was financed by taxing so many free people so much without their individual consent! It reminds me of the building of the pyramids in Ancient Egypt by slave labor. It is underappreciated that taking $180 billion in today's dollars from taxpayers for a project of almost purely symbolic value was actually an ideological concession to collectivism.

The pyramids weren't built by slave labor.

Huh. I was convinced they were. But you're probably right.

not slave labor, 'obligatory' labor.

Indeed. Think how much more could have been spent on the Vietnam War, running concurrently, which cost $738B in 2010 dollars according to page 4 of this congressional report:

Nol. Breaking out of a cell with dental floss is a hard problem and looks like people have solved it

4. If we are No longer able to solve big problems, we may just have to wait till thry become small problems.

For example in the 1960s getting to the moon was one of the biggest problems we could concieve, by 2030 it looks like it won't be that big a problem and will take a far smaller proportion of economic resources. This can be extrapolated to a whole set of problems, for example an Egyptian official or noble spent his whole life and a huge proportion of available resources to build a tiny rock cut tomb that could be built today as a weekend hobby. I just picked that example due to its ridiculousness.

Epic fail by the WaPo. How much revenue is less from the Romney plan depends on your 'baseline'. The Romney plan is meant to be revenue neutral w.r.t to revenue raised under the 2012 law. So rather than the ~450B in the calculator, the revenue loss is 247B.

What the Romney plan doesn't do is provide the extra revenue that would come from ending the Bush tax cuts.

"have we lost the ability to solve big problems?"
1) The government just made solving many of them illegal.
2) It became so cheap to solve many of them, that no one sees them as "big" any more.

Comments for this post are closed