News Bits

by Luke Muehlhauser on March 6, 2011 in News

This blog has content very similar to Common Sense Atheism.

I updated Ethics and Superintelligence again.

I will be posting on the Reading Yudkowky series more frequently for a while so I can get through them all.

From Twitter:

Previous post:

Next post:

{ 31 comments… read them below or add one }

Tom Morris March 6, 2011 at 4:43 am

The Wikipedia as a force for evil link: “In 2010 the number of editors declined to 50,000 while the number of articles in English alone had increased to ten million.” is followed by a link to a graph showing that the number of articles in English Wikipedia is just over 3.5 million.

And as for people keeping an eye on articles: personally, I usually have about 650 articles on my watchlist, as well as keeping an eye on deletion discussions for philosophy and atheism.

“This means that eventually society will have no way to determine if a Wikipedia article is telling the truth or not.”

Sure there is. If it has a little purple star thing, it’s a featured article. If it has a little green blob, it’s a good article. And then you check to see if there are footnotes to reliable sources. Those reliable sources are usually books, newspapers or websites deemed reliable (like academic or scientific websites).

The author of the linked page needs to learn more about Wikipedia: it’s a lot less scary than he imagines, even though there are people editing it to push their POV.

  (Quote)

JS Allen March 6, 2011 at 7:29 am

Steve Connor comes across as an insufferable douchebag in that interview. It’s a good thing he doesn’t actually do science.

  (Quote)

Sharkey March 6, 2011 at 7:47 am

@JS Allen: I was just thinking the same thing, except replacing Dyson for Connor. Dyson repeatedly retreats into the favoured responses of deniers: “I’m just asking questions”, “I don’t accept the premises of the argument, and refuse to answer why”, and “I don’t feel like talking anymore”.

Perhaps Dyson is more skeptical in other venues, but I only saw a stereotypical denier in that interview.

  (Quote)

Martin March 6, 2011 at 8:44 am

Wow. That Freeman Dyson interview. Dyson quickly paints himself into a corner and then has to get testy and bow out.

Steve Connor is 100% correct: climate “skeptics” have no consistent argument. One day the warming is not happening at all, then it is, but its the sun. Now its CO2 but it won’t be bad, and anyway, it’s cooling. It’s El Nino. It’s cooling. It’s warming but it’s the sun. It’s CO2, but it won’t be bad, but it’s cooling. The instrumental record is unreliable, but it shows cooling. But it’s the sun. It’s warming. And so on.

This is quite true. On Anthony Watts’ blog you can witness the self-contradictory posts from day to day. A legitimate question.

Dyson’s response? “Blah! I quit!”

Dyson is a perfect example of the strongest of the Dunning-Kruger effect. He’s so smart about physics that he (erroneously) believes himself to be smart about other sciences as well.

  (Quote)

Polymeron March 6, 2011 at 10:02 am

I don’t hold much esteem for “Take on it” style projects. When you ask a difficult question such as “Will the WikiLeaks cable leaks do more harm than good?”, you can’t simply expect to ask any old expert and receive a definitive, reliable answer. At most you’ll get an argument or two that you don’t have an immediate answer to.

No matter how many times you shout “let’s get this sorted!”

  (Quote)

JS Allen March 6, 2011 at 10:12 am

@Sharkey, @Martin – OK, good to know that smart people had exactly the opposite reaction to mine. I’m not a global warming skeptic myself, but I thought the exchange was surreal:

Dyson: “Experts are often wrong, and the computer models are bad”
Connor: “But you’re wrong, because experts say… And the computer models say…”
Dyson: “… oh, in that case, I’ll point out that experts are often wrong, and the computer models are bad”.

It was a total failure to communicate. Connor even said, “The scientists who handle these models point out that they can accurately match up the computer predictions to real climatic trends in the past”. …. Yes, he said that! If I were Dyson, I probably would have bowed out at that point, explaining, “You are literally too stupid to talk with”. It’s like Connor is either stupid, or being a troll. Since when is fit with past data proof of anything?

I’m sympathetic to Dyson’s general cautionary tone, since I saw the exact same thing happen in 2006-2007 when “heretics” like Bookstaber, Taleb, and Roubini were warning that the experts and computer models were wrong about the economy. I took all of my money out of the stock market right before the “experts” blew up the entire world economy with their “computer models”.

We’re still witnessing the ripple effect of human misery from that blind faith in experts and computer models. And the economy is something that humans invented and control. The economy is clownishly simple compared to the global climate. If we can’t even model the economy without screwing it up and destroying people’s lives, why would we place too much faith in our climate models? Our confidence intervals need to reflect the weaknesses of the model.

Personally, I think there is a divide between practitioners and the general public when it comes to trust of models. I have a good deal of professional experience creating complex models, including economic models. My professional experience predisposed me to pay attention to the “heretics” like Taleb and Bookstaber. Among my friends who have professional or academic experience modeling complex systems, none of them place a high confidence interval on climate models. That doesn’t mean we’re global warming deniers, but means that we don’t place a lot of faith in the models.

There was a recent experiment where people were told that a computer was going to drive the car for them. In reality, a hidden human drove the car, but the passenger thought it was computer driven. The vast majority of people got in the car and let the “computer” drive them. To me, that shows the level of stupid faith that the general public have in computers.

  (Quote)

Sharkey March 6, 2011 at 10:49 am

@JS Allen:

I don’t disagree with your skepticism towards computer modelling; indeed, climate modelling is a superset of economic modelling, considering that that climate change argument assumes that climate is influenced by economic factors, so any difficulties modelling economies will be compounded for climate modelling.

However, I do have a couple of nits to pick. For one, your Connor quote is taken out of context slightly. Fit to model doesn’t necessarily provide predictive power, but it can have explanatory power. In Connor’s quote, he implies that models that disregard CO2 heat trapping don’t track historical information, and adding that behaviour models historical trends. It’s true, it’s possible that CO2 only trapped heat between 1960 and 1990, and will change it’s behaviour from 2015 onwards, but that would need confidence intervals and explanations as well. On it’s own, it acts as support for the premise that CO2 traps heat, which Dyson disputed without explanation.

Secondly, a recent study on the predictive power of experts such as Taleb casts doubt on their abilities. Extreme yet rare predictions are made very frequently, yet we only take notice of those small subset of predictors that managed to be the correct “stopped clock”. Put simply: the fact that experts can be wrong doesn’t exempt Taleb nor Dyson from that very property.

Finally, the only fix for faulty models is independence. If multiple models tracking independent variables lead to consistent and converging results, then the results are worth considering. I leave it to the statisticians and climate scientists whether the set of current climate models have this property…

  (Quote)

Esteban R. (Formerly Steven R.) March 6, 2011 at 11:35 am

I’m actually much more sympathetic to the view of Wikipedia being used to misinform people. I recall looking through “List of Emo Bands” and seeing bands like “30 Seconds to Mars” on it. A bit surprised, I went to their page and saw that they were labeled “emo” and, even more surprising “post-hardcore”. Surely this was just a musically oblivious fan editing the genres, right? A check in the discussion page revealed that all the sources justifying those dubious genres came from “legitimate” sources, that is, two poorly written articles published on the internet by some minor entertainment sources.

Now, this all seems trivial but if you go to any major controversial topic, discussions where poorly written articles by misinformed people are passed off as legitimate sources happen all the time. If someone blogging about music and getting their stuff linked by an entertainment site meets Wikipedia criteria for an article, then imagine just how easy it is to make “valid” changes on important information.

Next, a look at philosophical arguments for the existence of God reveals the same pattern. Read the page on the Argument from beauty: http://en.wikipedia.org/wiki/Argument_from_beauty

The whole article reeks of individual research, poor research and someone using Wikipedia to further their Theistic Argument. The current version, mind you, is greatly edited and improved from when I first encountered it. It used to be written in a “Argument to back up the premises”-”Stupid objection raised by stupid materialists”-”My argument against it, but because I have a citation that kinda uses the same words I’m using and comes from a reliable source, it is legitimate” manner and greatly biased to saying the argument succeeds.

In fact, look at the “Truth, Beauty and Science” section. It’s just made up of a number of quotes that associate mathematics and science with beauty, which tells us nothing about the scientific process or mathematics, but is obviously used to try and make it seem as if scientists accept the argument. Finally, the criticisms section is a mess. It doesn’t understand the argument, but even then, the Theistic bias persists. We see a quote from McCabe and then “but this ignores the Taoist concept of polarity”. Where did that come from? It’s just something not even related to the argument at all and it’s just the same tone of “oh, this argument is great, all objections are stupid” that I noted the first time I read the Wikipedia article.

Depending on the day, the Cosmological or Teleological pages may have a decent overview of the arguments, may not even feature a criticisms section at all (or feature the weakest criticism and then a long, nonsensical rebuttal that somebody clearly thinks they’re an apologist) or may not even understand the argument at all.

Lastly, apparently some students came up with a way to track who is editing Wikipedia pages (I’ll see if I can find the link later) and when they got the results back, computer changes came from all sorts of places, including the CIA. Forgive my cynicism, but this seems really, really scary, because given the poor standards that Wikipedia has and how it’s so easy to brush aside any objections by shouting “original research!” that I can see how Wikipedia is a poor source of information. Granted, a critical reader will be able to spot it, but I wonder how many people would be impressed by the idea that because it’s Wikipedia, it must be correct.

  (Quote)

Esteban R. (Formerly Steven R.) March 6, 2011 at 11:41 am

Long post out of the way:

Luke:
I remember reading Krugman’s column earlier this week and just being amazed at the sort of corruption going on. So sad to see how people still believe union bargaining somehow leads to major state deficits.

Thankfully, The Daily Show had a hilarious–and very good–critique of “conservative” rhetoric: http://www.thedailyshow.com/watch/thu-march-3-2011/crisis-in-the-dairyland—for-richer-and-poorer—teachers-and-wall-street

Thanks for the links, some pretty interesting stuff here. Including a new music recommendation. Time to listen to the Dirty Three.

—-

Interesting exchange between JS Allen and Sharkey. I really am getting the impression Martin got though.

  (Quote)

stamati March 6, 2011 at 11:56 am

I thought Connor asked good questions, or at least they were questions I would like the answers to. I found it disappointing that Dyson bowed out instead of illustrating his stance. You guys know any good resources about this debate (if it even is one)?

  (Quote)

Martin March 6, 2011 at 11:59 am

JS Allen,

Dyson: “Experts are often wrong, and the computer models are bad”
Connor: “But you’re wrong, because experts say… And the computer models say…”
Dyson: “… oh, in that case, I’ll point out that experts are often wrong, and the computer models are bad”.

Appeal to authority is only a fallacy in certain cases. According to my logic textbook, appeal to authority is non-fallacious if used in the following way:

1. X holds that A is true
2. X is an authority on the subject
3. The consensus of the authorities agree with X
4. There’s a presumption that A is true

The fallacious form dumps premise 2 or 3. Appealing to Dyson’s opinion on climate would be omitting both premises. It’s true that authorities are sometimes wrong, but it isn’t common. If Dyson thinks they are all wrong, then he should publish something in a journal showing specifics. Journals could make a name for themselves overturning a paradigm, and would eagerly publish something that shows all climate scientists in the world are wrong.

And secondly, there is plenty of evidence outside the realm of climate models that back them up. There is an enhanced greenouse effect as measured by satellites, so we can now empirically detect the presence of global warming. More heat is coming in than going out. And the heat that is not going back out is of the same wavelength that CO2 is known to absorb.

Coupled with a quiet sun, and ever increasing CO2 and temperature, you get a damn good case for global warming.

Dyson doesn’t know what he is talking about, and his opinion is directly contrary to people who work with climate models all day every day and have devoted their entire lives to the subject. Any perceived douchiness on Connor’s part is probably part of the same frustration that us “Darwinists” feel when we have to deal with creationists and IDers. We’re sick of anti-science horseshit being spouted from people with an ideological opposition to the subject in question.

  (Quote)

Garren March 6, 2011 at 12:09 pm

Luke,

Since you liked my last couple of posts about hypothetical imperatives, I thought I’d let you know I just wrote another after stumbling across an enlightening quote in a (fantastic) philosophy of science textbook.

  (Quote)

strangebeasty March 6, 2011 at 12:11 pm

Did you ever post your interview with Russell Blackford anywhere?

  (Quote)

Luke Muehlhauser March 6, 2011 at 12:43 pm

strangebeasty,

No. The podcast is on hiatus for a bit.

  (Quote)

Paul March 6, 2011 at 1:42 pm

I posted this at ivorytowermetaphysics:

How math describes reality is less of a mystery if you view math as a human construct. Math need not be some window into ultimate reality, in a similar way that our perceptions are not a window into the ultimate reality of an object, but are merely a reflection of the object filtered through our senses. Redness, as a qualia, is not in the apple. When we perceive the redness of an apple, we are not glimpsing something essential about the ultimate nature of the apple, because that redness isn’t in the apple.

Similarly, math may describe aspect of the universe to a greater or lesser degree. When the degree is greater, it need not be anything more than a tool that we have constructed that works. It works exceedingly well, but there’s nothing demanding that it reflects something essentia

  (Quote)

JS Allen March 6, 2011 at 4:07 pm

Fit to model doesn’t necessarily provide predictive power, but it can have explanatory power. In Connor’s quote, he implies that models that disregard CO2 heat trapping don’t track historical information, and adding that behaviour models historical trends

Sure, I understood what he was trying to say. But you can see why a physicist would not be impressed. Physics and economics learned this game long ago — you can just add some more dimensions to your model until everything fits. I’ve done the same with my own models. But its far too easy to delude yourself with the math. The only way to keep yourself honest is to then use those models to make predictions against new data. If your model isn’t falsifiable, you’re screwed.

I can understand why this is an impossibility with climate data — to be sure the model is right, you would need to monitor for the next 100 years and see if new data fits the model. And if they’re right, we may not have that much time.

However, physicists would never put up with that. Most models end up being wrong, and you need experimental data to test and falsify them. The fact that climate “scientists” don’t have this is more than enough reason for an accomplished physicist to reject the claims without argument. Climate scientists may be right, and may even probably be right, but they didn’t arrive at that answer the right way. They are making a big huge guess.

I don’t think that many non-practitioners get this point, which is why they don’t see how stupid Connor’s question was. There are cases where we can’t set up experiments to falsify the model, so in those cases we sometimes randomly partition the data. We use half the data (for example) to train a model, and then we check the model against the other half to see how predictive it is. However, this falls prey to many of the flaws that led Daryl Bem to claim that he had proven Psi. I have seen this myself with genome and disease data, where I can make a model that exactly matches one partition of data, and predicts the other, yet is completely wrong. If you haven’t read the paper explaining why Bem was wrong, and what needs to be done about it, you should. IMO, climate science is on a similar footing. Just replace “psi” with “carbon”.

Secondly, a recent study on the predictive power of experts such as Taleb casts doubt on their abilities.

While I agree 100% with you, I hope you realize how ironic this sounds. Taleb has never claimed to be able to make predictions. In fact, he has made a career on claiming that he can’t make predictions — he has no idea what is going to happen, but if someone else claims to be able to predict, then he’s certain that they are wrong. He was among the first to publish research showing that the “experts” do worse than randomness.

It’s true that authorities are sometimes wrong, but it isn’t common.

When have the experts ever been right on something of this complexity and magnitude? We’ve already seen how they did with the economy, which is a simpler problem. And we know they can’t predict the weather out a week in advance. And note that we are talking about Dyson, who is an expert in modeling the physical world. Other accomplished physicists also point out that climate models are complete shit that no self-respecting scientist should trust too much. Lubos Motl is just one who comes to mind.

FWIW, I believe that the experts are right that A) the climate is warming gradually, and B) humans are contributing to some portion of that warming. But everything after that gets into sci-fi territory.

Dyson doesn’t know what he is talking about, and his opinion is directly contrary to people who work with climate models all day every day and have devoted their entire lives to the subject. Any perceived douchiness on Connor’s part is probably part of the same frustration that us “Darwinists” feel when we have to deal with creationists and IDers. We’re sick of anti-science horseshit being spouted from people with an ideological opposition to the subject in question

Well, if Dyson doesn’t know what he’s talking about, I’m certain that you and Connor know even less. At least Dyson has a lifetime of experience modeling complex systems. It may be tempting to see idealogical boogeymen behind every corner, but that’s not a very scientific way to conduct your life. If Dyson is a boogeyman, and the other physicists who point out the problems are boogeymen, what does that make Connor? Do you give him a free pass simply because he agrees with your ideology?

  (Quote)

Sharkey March 6, 2011 at 4:43 pm

@JS Allen:

I think we are mostly in agreement, with perhaps slightly different interpretations of the utility of the models (and I say this not as a climate scientist, just a computer scientist).

As always, a few nits:

- Taleb does make a career of predicting, both publicly (most notably about the crash), and privately as a consultant and a hedge fund advisor.

- Physics and economics are two scientific disciplines with both the ability and financial motivation to perform predictive experiments on timescales applicable to currently living humans. Connor’s answers may not impress a physicist, but if that was the golden standard for scientific rigor, much of what is currently considered interesting science would be dropped. And why choose physics? Why not go all the way to the only rigorous discipline: I speak, of course, of mathematics :)

I agree that computer models can only tell you true things about the model; any application to the real world is tempered by how closely the model’s axioms reflect reality.

  (Quote)

Martin March 6, 2011 at 4:47 pm

And we know they can’t predict the weather out a week in advance.

Because the weather is like a single roll of a set of loaded dice. You will never be able to predict each individual roll, but you will be able to predict that 7s will come up more often than not. Same with climate. Climate is weather averaged over at least 30 years. Fully predictable, because there are only a handful of factors that can influence it: the sun, greenhouse gases, ocean currents, Milankovitch cycles, and tectonic activity. If you learn the trend of all those factors, you can make predictions about what the long term trend in climate will do, even if you can’t predict the day-to-day weather.

But everything after that gets into sci-fi territory.

Science fiction of the type that is published in Science and Nature and other first-class science journals?

Well, if Dyson doesn’t know what he’s talking about, I’m certain that you and Connor know even less.

I would submit that I am freeer of Dunning-Kruger than Dyson is, because I don’t know shit about computer models. Therefore, I’m not likely to over estimate my competence at evaluating them to such a degree that I assume I’m better than, again, people who have devoted their entire lives to climate models. The problem with Dyson is that he does indeed understand computer models, but not climate models specifically, which are quite different. And since people who work have worked with them their entire lives say they work, and there are thousands of them that say this and have peer-reviewed articles in journals to back this up, I’m more inclined to believe them than Dyson, who doesn’t work with climate models and has no published work crticisicing them and is just a single individual.

Not to say that this is absolute Truth, but…

The probability that thousands of climate scientists who work with climate models every day at NASA and NOAA and publish their results in Science and Nature are correct > The probability that a retired physicist who never works with climate models and never publishes is correct

I mean, there are mega fucktons of peer-reviewed tests of climate models: http://www.skepticalscience.com/climate-models-intermediate.htm

If Dyson knows something these people don’t, why doesn’t he publish?

It’s all just generalities, and if pressed for specifics, as he is here, they just run away in anger and start yelling about “alarmists” and “propaganda.”

If Dyson is a boogeyman, and the other physicists who point out the problems are boogeymen, what does that make Connor? Do you give him a free pass simply because he agrees with your ideology?

He gets a free pass because he agrees with almost every climate scientist on the planet, who are the proper experts to appeal to in this particular issue.

  (Quote)

Reidish March 6, 2011 at 5:00 pm

Sharkey,

Taleb does make a career of predicting, both publicly (most notably about the crash), and privately as a consultant and a hedge fund advisor.

This really isn’t quite the case. He simply recognizes that the models many others are using severely misprice risk. The misapplication of Gaussian distributions has been the downfall of many a-leveraged market participants. He arbitrages their misunderstanding of their own models by buying cheap puts, and just waits for the fat tails.

  (Quote)

JS Allen March 6, 2011 at 5:07 pm

Martin,

So you’re relying on an appeal to the authority of a small cabal of self-selecting “experts” and a blind faith in massively complex computer models that you have no hope of understanding. At least you’re being honest. I don’t mind people doing that, as long as they’re honest about it.

And say what you want about climate change being simple and predictable. Until you make predictions that are better than Darly Bem’s, you’re just baldly asserting. I don’t think you have any clue, and even Connor doesn’t claim that much certainty. He basically evokes Pascal’s wager. “The models may be wrong, but what if they’re right?” It’s a big-ass guess and a gamble.

One other point about people who have “dedicated their lives to climate science”: You make that sound as if they are making some great sacrifice in the name of mankind. It’s like saying that Dick Cheney dedicated his life to protecting the security of all Americans, so we can trust him. More accurate would be to say that the climate “scientists’” financial security depends on politicians believing that there is a problem. And you make it sound like they evaluated other sciences, and then chose climate science because they were passionate about dedicating themselves to a good cause. I somehow doubt that many of them would have turned down the opportunity to be physicists if they had qualified. They certainly seem to have physics envy, and try to put a physics-like veneer of respectability on their models.

  (Quote)

Sharkey March 6, 2011 at 5:44 pm

Reidish, JS Allen:

You’re right, rather than Taleb I was thinking of Roubini, now that I go back through my articles. Mea culpa. Although Taleb does make predictions (in 2010 he incorrectly recommended shorting Treasuries, and he has a set of predictions for 2036), he’s much quieter than the other roster of economic experts.

  (Quote)

Garren March 6, 2011 at 6:36 pm

I’m quite interested in this topic of how to tell when good science is being done. Does anyone have suggested books or talks about it?

  (Quote)

Ajay March 6, 2011 at 7:26 pm

Garren – I’m no expert on the philosophy of science, but a great introductory book I’ve read is “Theory and Reality: An Introduction to the Philosophy of Science ” by Peter Godfrey-Smith, a philosophy professor at Harvard. It’s very basic and comprehensive, written without jargon.

  (Quote)

Luke Muehlhauser March 6, 2011 at 8:27 pm

Good rec, AJay.

  (Quote)

Michael March 7, 2011 at 3:45 am

This one is my favourite link of the week.
What an inspiring man.
“I will die to defend their rights.”
http://www.bbc.co.uk/news/world-south-asia-12617562

  (Quote)

juhou March 7, 2011 at 11:02 am

On the Freeman Dyson link. Here is an hour interview from this monday of Freeman Dyson related to climate change: http://www.econtalk.org/archives/2011/03/dyson_on_heresy.html

  (Quote)

Luke Muehlhauser March 7, 2011 at 3:04 pm

juhou,

Thanks!

  (Quote)

AndrewR March 7, 2011 at 7:34 pm

As soon as a philosopher begins talking about mental states as though they are fundamental or uniquely important things and not convenient abstractions for talking about particles bopping around in a particular way, you should know you’re in trouble
- http://www.wrongbot.com

Amen to that! It applies to that “Logic exists therefore God exists” bollocks as well.

  (Quote)

Michael March 8, 2011 at 5:13 am

Luke, I have a question.
I can’t remember where it was, but you had a list of people who you thought were the most influential ever, or something like that.
One was a medieval islamic philosopher who founded the philosophy of science I think, and another guy had developed some kind of agricultural technique or technology.
Do you know the names of those guys and/or where that list is now?
Thanks a lot for your help!

  (Quote)

Luke Muehlhauser March 8, 2011 at 7:16 pm

Michael,

I think you’re talking about this. But note that I no longer defend the list.

  (Quote)

Michael Baldwin March 9, 2011 at 4:30 am

Exactly what I was looking for, thanks.

  (Quote)

Leave a Comment