Assorted links


>>> #5 . "Is quantum computing basically upon us?" <<<

Absolutely not! That article had lots of hype. In the Quantum Computing community D-Wave's claims are considered with great skepticism.

D-Wave's magical black-box may be neat, but probably isn't a true Q-Computer. The largest problems that "true" Quantum Computers have solved so far are something like prime factoring the number 24 or so.

QC may or may not end up being revolutionary but it definitely isn't here yet!

Quantum computing has been right around the corner for so long ive kind of lost track. Its the IPv6 of the chip world. Funny, the predictions of quantum computing performance always seem to increase with each new burst of hype.

At least IPV6 exists.

A propos the debt graph by our Grumpy Economist: "Are the 'Austrians' right to worry about the long term? Well, it's surprisingly hard to say. The only thing relatively certain in the CBO or CBPP hypotheses is the demography".

I would add that John Mauldin had made similar comparisons a while back. Over time, compounding of anything - growth rates, interest payments etc - looks staggering. That's compounding, nothing special.

And I doubt the solutions the Grumpy Economist has in mind would do much for 'growth' so...

Re: Bush's paintings

The comments are revealing. Just as some on the right hate Obama with a religious fervour so many on the left hate Bush in the same way. If GWB had studied at an art school or had pretensions re: being an artisan, let alone a contributor to "The Art World", perhaps one could mock him.

He's a fairly representational kind of painter, broadly in the impressionist tradition. Let me ask you: what would have been a suitable genre for GWB? What would be said if he'd tried to do Rothko-esque monochromes? What if his art consisted in hurling elephant dung? Peeing into jars? Sharks in formaldehyde? Meat dress? Pile of rubble? Damned if you, George and damned if you don't.

Agreed, the comments to the story tell you more than the story does.

Bush is a good painter. Pretty charming and whimsical. The one of his legs in the bath is kind of touching in some way. Still a horrible president.

True. Were people as deranged in the 20th century? I remember people not liking Clinton and Bush, but I don't remember the same level of acrimony.

Should'a heard what they said about LBJ.

Not really having the Internet and well-developed blogs around meant you didn't really see the crazy side of people (different from crazy people; ) before the 21st century.

The internet was well established during the Clinton years.

"He says he has support from a group of retired professional NHL players he knows through his father"

Oh, well, retired athletes NEVER make poor investment decisions, so it must be a good idea.

I like how Canadian stories like #1 tend to tie back to hockey regardless of the subject matter. If #2 was about Brian Mulroney's paintings, they would probably be paintings of pond hockey. And dogs watching pond hockey.

The general consensus among physicist, as confirmed by my time at the recent APS meeting, is that D-Wave is not on track to produce a useful quantum computation any time soon. Their preferred substrate is not the most favored and, more importantly, they simply haven't shown that they are actually doing *anything* truly quantum. They are good at hype though.

It's funny, Tetlock's study might be one of the most important ever done, and yet we continue to pay undue heed to expert predictions.

First, Tetlock's study was largely about *political* prediction; it's not a blanket permission to ignore expert opinion. Second, you might reasonably ignore the physicists if you had an outside view analysis to fall back on. It's pretty hard to find a good reference class, though, and things like fusion energy immediately come to mind. Certainly, the self-described progress of a private company in the hunt for government dollars is not a more reliable source.

Hedgehog prediction strategies may not be very good at making successful predictions, but they're extremely good at:

a. Helping make sense of a mess of datapoints by weaving them into a story.

b. Convincing people that you know what you're talking to, probably largely because of (a).

Without setting up a pretty artificial experimental structure, it's hard to really gauge how accurate people's predictions are, and people are probably often fooling themselves about how accurate their predictions are--probably also with a pro-hedgehog bias.

So one natural guess is that absent some such artificial structure for really keeping track of the quality of predictions, hedgehog strategies are massively overproduced by public thinkers. That's what they get rewarded for in the far more common case that they're in school trying to remember a lot of details, and later when they're in the business of convincing other people that they understand some complicated, poorly-understood thing.

What I'd underscore in the article is the idea that the same forecaster should be forced to switch back and forth between being a hedgehog and a fox (or taking "the inside view and the outside view"). I work on a rather hedgehog-intensive forecast, but I love it when I get to (and am asked to) voice my fox views. At first I thought I was trying to win over the story to my little bits of insight, but that doesn't generally happen. Fox views are not well suited to a coherent story and no policy maker is going to start doing stuff without a story. The fox gets to pull on the edges and just be a little sand in the wheels. That said, I think am a more effective forecaster doing both, spending two-thirds of my time as a hedgehog and one-third as a fox....and above all paying attention to when each role is most productive. One thing missing from the article is the idea that for better or for worse, forecasts are not always about the forecast, but also about how it will be used.

Is Tetlock himself a fox or a hedgehog? In any case, I like the idea of a mixed fox/hedeghog strategy ... I don't like Brooks hero-worship of Tetlock

I wonder what Watson would say about all of this?

Has Tetlock played his prediction games with Watson as a player?

No, we are using only human intelligence so far, but we are experimenting with ways to scrape news and generate predictions that way. We are several years away from the time where Watson, or any other algorithm can match skillful and diligent human forecasters.

Argh, I'm completely stuck on that in #2 the people at Gawker think a very obvious donkey (it's his best painting of the lot, if you ask me) is a horse. It could possibly be a mule, but it is absolutely not a horse. I swung down to the comments assuming they'd get pinged on it and no, not really, just a few people.

How did they all miss that? How does a whole staff of people and an army of readers think that's what a horse looks like?

> How does a whole staff of people and an army of readers think that’s what a horse looks like?

We've got a couple of plausible explanations:

1. They all live in NYC and have never seen a mule nor donkey.

2. They were so busy formulating their "responses" that they barely bothered looking at the paintings

The prediction project is neat, but predicting binary results are extremely limited. That throws away SO much data about who is a good forecaster and who is a bad forecaster (if a forecaster is only right or wrong, rather than close to what actually happened or far from it).

Besides, most of the outcomes that we care about aren't binary, they're continuous (unemployment rate, GDP growth, deaths in civil war, deaths to flu, etc).

Tetlock, man up! Quit thinking binary! Think continuously!

1. Calling the top.

I feel bad for Bush that those paintings are out there. Despite a few feeble attempts to defend his talent, they are awful. But they're awful in the way that abject beginners make awful paintings, and he probably had some pleasure in making them. Now that he's been outed and mocked, he'll probably quit and go back to clearing brush for a hobby. A better strategy than making fun of him would have been to co-opt him in a push to bring more arts education back into public schools.

Haters are going to hate. Bush is aware of that and he's got pretty thick skin at this point.

Not hating, just stating a fact. They'd be awful if they were done by [fill in the blank with anyone else you admire]. The paintings have every hallmark of an untutored beginning painter. Nothing wrong with that, since that's what he probably is. And I'm not sure what Bush's awareness or the thickness of his skin has to do with anything.

Not hating, just stating a fact. No you stated an opinion, not a fact. And I wasn't necessarily saying your comments were hateful, but that plenty of the comments were hateful. And you did say, he’ll probably quit and go back to clearing brush for a hobby. My response is that he's heard it all, has thick skin, and probably won't change his mind about a few haters.

During graduate school, one of my best professors said the only thing he was going to teach us was how to read.

Reading is hard.

Sorry, you weren't very clear. I read your response to suggest that I was being a hater for pointing out that Bush's paintings are awful. And yes, you're correct, what I stated was an opinion, in much the same way that a professional ballet dancer and dance instructor might go to a high-school dance recital and state their opinion that the dancing was awful and showed all the hallmarks of untutored beginners.

What I should have said was, "Not hating, just giving a professional opinion from having done art for a living, and having taught art, and having been subjected to the artwork of countless self taught amateurs."

Comments for this post are closed