Disqus - Latest Comments for clumma

Re: “In 2030, the share of world energy produced by solar will be < 2%, conditional on real gross world product (in 2015 USD at market exchange rates) being not less than $118 trillion. End date as soo...

Carl Lumma — Fri, 08 Jul 2022 18:47:52 -0000

Sorry, just seeing your reply now. $200 is fine. However I want to change the prediction somewhat. I will contact longbets about this.

Re: “In 2030, the share of world energy produced by solar will be < 2%, conditional on real gross world product (in 2015 USD at market exchange rates) being not less than $118 trillion. End date as soo...

Carl Lumma — Fri, 17 Jun 2022 17:03:28 -0000

I got an e-mail from longbets with a challenge from you. Thanks! I can reply to longbets to accept, but I'm unsure how to continue the discussion, other than to comment here.

Global GDP in 2020 was about 85T in current dollars. Hitting the high end of your range in 2030 at 150T current dollars would be ~ 5.8%/year nominal growth. That could easily contain the real growth rate I suggested (not sure how you got 200T).

But it could also contain 0% real growth and 5.8% inflation, and that wouldn't violate my thesis. So I can't accept the challange. It has to be based on real growth.

I can go down to 115T 2015USD, which is the least squares fit since 2000.

Re: “In 2030, the share of world energy produced by solar will be < 2%, conditional on real gross world product (in 2015 USD at market exchange rates) being not less than $118 trillion. End date as soo...

Carl Lumma — Fri, 17 Jun 2022 13:49:43 -0000

The $118T was based on a least-squares fit of RGWP 1970-2020. The actual fit is $119T - I subtracted a trillion for good measure. I'm willing to consider other terms though.

Re: “In 2030, the share of world energy produced by solar will be < 2%, conditional on real gross world product (in 2015 USD at market exchange rates) being not less than $118 trillion. End date as soo...

Carl Lumma — Fri, 10 Jun 2022 14:37:31 -0000

Energy production from non-thermal sources (such as solar PV) will be converted to thermal equivalence by dividing by 0.4.

At no time will the energy share supplied by solar exceed that supplied by fission or methane (whichever is larger) without dramatic reduction of the world economic growth rate of ~3.1%/year.

Re: The Eighty Five Percent Rule for optimal learning

Carl Lumma — Sun, 10 Nov 2019 11:01:54 -0000

The portion 15% seems to crop up suspiciously in optimization contexts... this was noted by Gell-Mann in The Quark and the Jaguar. It's roughly the portion of false warnings sounded by certain tropical
birds to gain uncontested access to food. Gell-Mann speculates that it
is close to 1/(2pi)...

Re: Information Processing: Kip Thorne on Caltech and Black Holes

Carl Lumma — Sat, 02 Mar 2019 20:07:04 -0000

Detecting gravity waves is amazing but doesn't require building a machine that can dissipate 150 GW for several minutes with humans inside of it. Our civilization routinely does the incredible on the margins but seems to lack the free energy for large projects. It's consistent with a high-technology civilization living on a shallow entropy gradient.

Re: Information Processing: Trading Blows: The US-China Trade War (Yukon Huang)

Carl Lumma — Thu, 06 Dec 2018 17:14:27 -0000

The "structural consequences" I mentioned are well within the purview of orthodox economics. Papers have probably been published arguing they outweigh the trade benefits, as well as papers arguing the reverse. It's a natural philosophy, in the worst sense, so no progress is made by such publications. But I didn't have to mention the orthodoxy issue, except that it is always a good idea to mention it on general principle.

Munger's comments are OK but don't quite get to the point. Ricardo's 2-country, 2-good model of 1817 doesn't "prove" anything, and two centuries later there is no consensus how to extend it to a more realistic setting, let alone prove anything with it. It's in the first chapter of every undergrad textbook but if you do a literature search, it's hardly been studied.

In fact, the "comparative advantage" is just the advantage of specialization, which doesn't have to involve international borders. Detroit can make cars; Los Angeles movies; Pittsburgh steel; and the domestic economy can be both self-sufficient and grow faster than the global economy is growing today. That happened so we know it's true.

The heterodox part is important because you'll always overestimate the benefit of trade if you believe it's the cause of industrial growth. But even in Ricardo's treatment, specialization confers an O(1) factor in the output level and hence cannot explain explosive per-capita growth since 1800, or its near total absence prior to 1800 despite millennia of increasing global trade and specialization.

More:
https://twitter.com/clumma/...
https://twitter.com/clumma/...

Re: Information Processing: Trading Blows: The US-China Trade War (Yukon Huang)

Carl Lumma — Wed, 05 Dec 2018 13:59:31 -0000

Now let's take an example. In the video, the iPhone was given. I'm not sure it's a canonical example of the "trade war", but let's run with it. The speaker has a PhD in economics from Princeton, which means he plausibly knows less about economics than when he graduated high school. I worked at Apple while the iPhone was being invented, and on every release through the 3GS, and the first iPad, and many other products. My job involved managing contract manufacturers in China (and Taiwan, Korea, Japan, and the US). I wrote software used by contract manufacturers, and audited their data. I wrote software used internally by Apple engineers to manage the same products and parts. I also studied the global supply chain, and worked on energy intensity and materials studies for all Apple parts.

When I left Apple, iPhone and iPod assembly employed on the order of 100,000 people in China. Apple had 30,000 employees in total (including retail personnel). No one knows the impact on US unemployment, but a number of additional unemployed > 30,000 can't be ruled out. The US lost 1,000,000 manufacturing jobs from Jan 2003 until the start of the recession.

How did the US pay these unemployment claims, and the resulting welfare, medical, and criminal justice costs? We borrowed the money from China. And then we took the interest payments out of my paycheck.

Does this sound like a good deal for the US? No one knows. I personally think it's unlikely, but it's important to stress that no one has any way to measure this, and that people like the speaker in the video are charlatans for implying they do.

Re: Information Processing: Trading Blows: The US-China Trade War (Yukon Huang)

Carl Lumma — Wed, 05 Dec 2018 12:50:04 -0000

First 9 minutes show a typical 1-D view of trade, which disregards the structural consequences of foreign-held debt, domestic unemployment, and loss of know-how. At the same time, it implicitly adopts the (incorrect) view of orthodox economics that trade is responsible for economic growth. At the global macro level, trade is asymptotically zero-sum. At the national accounts level, though, it can be tremendously beneficial to net exporters like China. And it continues to be so, despite the bizarre denial that trade is "only" 40% of China's economy (a huge number) and that it has generated negative growth recently (whatever that means). Finally, the association of Monday's stock market movement with the news from Argentina is completely fallacious.

Re: Information Processing: Kosen Judo and the origins of MMA

Carl Lumma — Thu, 08 Mar 2018 22:51:02 -0000

Rickson Gracie won Vale Tudo Japan twice in the mid '90s, and the event itself was from Brazil. This didn't create awareness of BJJ?

Re: 23andMe At ASHG 2017

Carl Lumma — Wed, 11 Oct 2017 12:48:03 -0000

When will papers/posters be available?

Re: Wigner recollections

Carl Lumma — Thu, 25 May 2017 17:50:01 -0000

Between this book and that of vN's daughter,[1] we get a good picture of the man -- I think a better one than a biographer could produce.

[1] https://www.amazon.com/dp/0...
Note that I don't recommend this book except for those willing to infer details about vN from a somewhat unreliable source while weeding through 67% unrelated stuff.

Re: Gallery: the cool and strange at Germany’s huge Musikmesse show

Carl Lumma — Sun, 07 May 2017 14:10:55 -0000

What is messe16-12.jpg!?

Re: Calculating Growth Rates | Dietrich Vollrath

Carl Lumma — Tue, 14 Feb 2017 02:14:39 -0000

> Our theories could be completely wrong

Unfortunately that does not seem to be true.

Re: More on Decomposing US Productivity Growth

Carl Lumma — Tue, 14 Feb 2017 01:45:19 -0000

Decomposition of GDP would look like PCA, not subtraction.

Re: Comparing Forecasts – PredictWise

Carl Lumma — Tue, 01 Nov 2016 11:54:29 -0000

Have you looked at Granger causality? I guess the problem is, it doesn't work if both forecasts are responding to the 'true' probability with different lags. There may be a way around that with VAR. . .

Re: Comparing Forecasts – PredictWise

Carl Lumma — Fri, 28 Oct 2016 13:51:29 -0000

Is the first method symmetric? I.e. what happens if PredictWise forecasts FiveThirtyEight?

Re: Guns, Massacres, and Ephemeral Shifts in Public Policy Positions – PredictWise

Carl Lumma — Mon, 22 Aug 2016 13:17:33 -0000

The trend among independents doesn't look very convincing. Are you basing it on just four pre-masacre samples?

Re: Prediction Market Ineffeciency – PredictWise

Carl Lumma — Fri, 19 Aug 2016 22:09:31 -0000

Excellent article.

What is the total volume on the State markets vs. the volume on the topline market(s)?

Re: Something’s Odd About the Political Betting Markets – PredictWise

Carl Lumma — Tue, 12 Jul 2016 14:32:22 -0000

Thanks, this is an important hypothesis. Would be good to check it quantitatively. Have you looked at prediction market volatility over time? Another idea: the beta between market probabilities and a series constructed from relevant polls. If people are increasingly discounting outside information, this beta should be falling.

Re: Brexit: leave voters relatively unfazed by currency collapse – PredictWise

Carl Lumma — Fri, 01 Jul 2016 23:27:20 -0000

I like the way you connect 3 data points with straight lines.

Did you see any such linearity within the individual (24 hour?) polling periods?

How many people were polled?

Re: Brexit: What Happened? – PredictWise

Carl Lumma — Mon, 27 Jun 2016 02:09:56 -0000

"At William Hill, 68% of the money ... was on 'remain', but 69% of the individual bets were placed on 'leave'."

https://twitter.com/PSMcCle...

Re: Brexit: What Happened? – PredictWise

Carl Lumma — Sun, 26 Jun 2016 13:30:58 -0000

Not at all. I said I've only seen two cases I'd call failure. That's fantastic performance, and of course I agree markets are currently the best prediction tools we have. It doesn't mean they're perfect. My informed opinion is that markets were wrong here, in a sense that could be made precise.

Re: Brexit: What Happened? – PredictWise

Carl Lumma — Sun, 26 Jun 2016 13:26:30 -0000

I've traded on markets you aggregate, quite deliberately as a Keynesian beauty contest. You just get out before the event occurs. I've been wondering about a mechanism to prevent it. Running a series of markets, say each a week long, which get frozen until the event is one idea I had. (Of course it does provide liquidity so it isn't all bad, and I'm not necessarily blaming it in the case of brexit here.)

Re: Brexit: What Happened? – PredictWise

Carl Lumma — Sun, 26 Jun 2016 02:21:02 -0000

Remarkably similar apologetic by Servan-Schreiber

https://blog.hypermind.com/...

One thing is certain: any criticism of prediction markets will be met with a lecture on calibration.

It's obvious something happened here. It should be studied.