Tuesday, November 25, 2014

Kuperberg and a physician vs a thinking smoker

Scott Aaronson has reposted a "parable" that mathematician Greg Kuperberg posted on his Facebook profile. Kuperberg wants to repeat the usual story that the climate alarmists' recommendations are just like the physicians' advises to their patients.

In the parable, a doctor wants her patient to stop smoking but she's unable to answer any questions (clearly meant to be quantitative questions) about the degree to which smoking is harmful. Surprisingly, although Aaronson's is a far left blog addressed to the Cambridge MA Lumpencafé, most commenters seem to realize that
  1. the certainty about the harmful impact of smoking is vastly higher than the certainty about any harmful impacts of CO2 in the atmosphere
  2. most readers seem to be "climate alarm solution skeptics", i.e. they don't think there exists a good rational basis or justification to "fight CO2-caused climate change"
  3. most of the readers actually seem to agree with the patient-smoker, anyway.
This discussion has lots of aspects – whether the particular technical claims about smoking and/or CO2 are right; what the impact actually is; how to behave in the typical situation when the scientific conclusions are statistical in character – and uncertain, too; what is legitimately included in people's and nation's decision making; whether experts' opinion is always the superior one; whether physicians and climate policy experts may be considered full-fledged scientists, and many others.

A reply addressing all the "dimensions" of this debate may look chaotic. Like the discussion at Aaronson's blog, it may look somewhat intelligent at the same moment – but it will hardly be too original or groundbreaking.

Just to be sure, I am convinced that science works in principle – even when it comes to the understanding of physiology, health, or climate processes. I think that most of the real-world science that makes it to textbooks at some level is right. I think that medicine works and physicians are generally close to science even though they usually don't do any research themselves.

At the same moment, I think that some professionals – especially physicians – may be imperfect and just mindlessly repeating something they have heard somewhere. Doctors have nevertheless helped me – and, even more often, people around me – in several or many situations. Sometimes, they were unable to diagnose a problem. Sometimes, they diagnosed a problem that was a non-problem. Sometimes, they would be trying to solve something that wasn't really a problem. But if one takes the union, the physicians undoubtedly improve the length and quality of our lives. That's not the case for paranormal healers. I am confident that homeopathy and most of the traditional schools of medicine fail to work. Some herbalist schools are partly OK. Various ancient civilizations may have known that certain essential oils are healthy, and so on. I don't want to go into technicalities. Their method to learn about these recommendations wasn't quite the scientific method so the percentage of "good ideas" is lower than it is in science-based medicine.

In the climate debate, as you know, the greenhouse effect exists in principle. The greenhouse effect caused by H2O, water vapor, is important enough: it makes the Earth's surface warmer by about 30 °C. The bare greenhouse effect caused by CO2 is just one among many additional drivers that are much weaker than the previously mentioned effect of H2O but that matter and as such a term, it could very well be neglected. Or not. If the feedbacks happen to be positive and large enough, which seems very unlikely at this moment, the warming due to CO2 itself may become visible or obvious in the 21st century. It's still not a problem and some warming – and especially the growth of CO2 itself, as a food for plants – is extremely likely to bring more benefits than losses. And even if it happened that the overall sign would become negative, perhaps sometime in 2150 if people still burn lots of CO2, it's very likely that the price of cutting the CO2 emissions would be much higher than the mitigated losses, so it would be silly to "fight the climate change", anyway. Some extraordinary discoveries or developments could in principle change my mind about any of these questions – it just seems extremely unlikely at this moment that something like that will take place.

Fine. Even though the case against smoking is much clearer and less corrupt than the case against the CO2 emissions justified by the climate alarm, it displays some of the same possible problems. For this reason, I think it is fair to use this "parable". Even in this parable where the "expert side" stands on a much firmer ground and even though I am a non-smoker (and I would surely be a combative non-smoker e.g. when I was a kid exposed to my mother's smoking), I am with those who say that the smoker's attitude in his debate with the physician was more intelligent than the doctor's.

The patient is right that the smoking-shorter_life link isn't "rock solid" and there are way too many exceptions. The link may only be detected statistically. And the patient is right that such statistical links are never 100% certain and don't prove a particular causal mechanism. Correlation is not causation. On the other hand, there is strong enough evidence that the causation statistically exists, as the doctor correctly mentions.

However, the doctor wasn't able to offer any numbers to the smoker. I will do it for you momentarily. More importantly, the doctor was a Fachidiot who considers the recommendations she hears from fellow doctors as a holy word – and seems to completely overlook the human dimension of the problem, especially the fact that people have some positive reasons to smoke, reasons she is completely overlooking. (In the same way, people may have very good reasons to veto a surgery or some treatment, too.) She looks at the patient as if she were looking at a collection of tissues and the only goal were to keep these tissues alive for a maximum amount of time. We sometimes hear that doctors are inhuman and look at their patients as if they were inanimate objects; however, it's rarely admitted that the physicians' recommendations "you have to stop A, B, C" are important examples of this inhuman attitude!

According to a 2000 wellness letter published by UC Berkeley,
Every cigarette a man smokes reduces his life by 11 minutes. Each carton of cigarettes thus represents a day and a half of lost life. Every year a man smokes a pack a day, he shortens his life by almost 2 months.
These are cute numbers. They're probably not "quite" exact or repeatable or constant but I think that if the correct number is 7 minutes or 15 minutes, it doesn't really change much about the underlying decisions because smoking isn't a precision science.

Let's think about the numbers. You don't know how long your life is going to be. If you consider yourself an average person in a nation, or an average man or woman, you may find some number, like 75 years or 80 years. The uncertainty or a standard deviation of the life expectancy is comparable to 10 years. Most TRF readers are intelligent men above 80 years of age ;-) – which also implies that they already know that their life is going to be longer than 80 years. Younger people sometimes think that older people won't live for too long – but these younger people often overlook the fact that the older people actually have a higher life expectancy than the younger ones. Those are paradoxes, aren't they?

Now, you may feel the urge to smoke. Let's assume that you need 11 minutes to smoke the cigarette. The smell of the cigarette is pleasant, sexy, and elegant for you. I have no idea why ;-) but I have seen enough to understand that many people are being attracted by totally different things and smells than your humble correspondent. Also, the cigarette shows you how powerful you are – you are free to ruin your body and ignore all the buzz in the media and in the hospitals. This freedom is great, too. I am partly joking – I enjoy my freedom not to smoke much more than the freedom to do harmful things, but I am still able to understand that different people may want to use their freedom differently.

When you live, you are spending the minutes that were reserved for your life (although the precise decision how much has been reserved isn't being made in advance, due to the free-will theorem in quantum mechanics). When you are smoking, you are spending your time at a doubled rate because those 11 minutes for the cigarette remove 11 minutes from your life because you got older and 11 more minutes because your life expectancy got shorter.

Life may be a fun process to be enjoyed. And those who like cigarettes probably think that the 11 minutes they spend with their cigarettes are "more than two times better quality time" they may be spending on Earth. If you like to live and you think that the quality of your life "less than doubles" during the 11 minutes you are spending with the cigarette, you should just stop smoking! This is a more specific lesson I would tell my patient who would ask about smoking.

This criterion is arguably rather useful because I think that you will find people on both sides. There will be people who don't really like smoking much and many of them won't think that the cigarette adds a whole 100% to the quality of the time they are spending with this particular man's best friend. There will be enthusiastic smokers who will think that those 11 minutes with a cigarette bring them much more than 2 times the quality-of-life than 11 minutes without a cigarette.

And there may actually be smokers who will find out that this shortening of their life is really too much and they may decide that it's stupid to smoke. And there may be non-smokers who will realize that the harmful impact of one cigarette is much smaller than they were led to believe and some of them may actually start smoking after they read this blog post because they always knew that cigarettes were great and they were just being brainwashed to think otherwise. I wish them to enjoy their shortened but improved life. ;-)

One may also discuss the harmful effects of smoking "macroscopically". If you smoke one pack a day for a decade (more precisely, 12 years), you will shorten your life expectancy by 2 years. Even if you intensely smoke for three decades, the effect on your life expectancy, while substantial, will remain below the standard deviation – the uncertainty about the length of your life that you really can't get rid of. In this sense, the question whether you smoke much can't become the "decisive" issue that determines the life expectancy of an average human.

Aside from the personal pleasure, "improvements of the economy" and "extra tax revenues" are quoted as additional positive side effects of smoking. Well, I think that this argumentation is largely irrational. If the people weren't spending their money for the cigarettes, they would be spending it on something else and this "something else" would bring about the same GDP, employment, and tax revenues.

The physician is likely to know much more about the contamination and deterioration of the tissues that may be caused by the cigarettes. Every patient may in principle learn these things as well; but he is very unlikely to beat a good enough doctor. However, even a patient with average intelligence should be able to understand the main number above – the shortening of the life by 11 minutes resulting from one cigarette. You don't need to be an expert to "interpret" this number. Of course that it must ultimately be the smoker's decision whether he wants to smoke another cigarette or not. And if his doctor were able to tell him at least one number – she wasn't – he would have everything he needs for an informed decision! I think that the whole conversation in Kuperberg's parable suggests that the smoker is much more rational when it comes to the "practical application" of the knowledge that the physicians have accumulated. In comparison, the doctor looks like a lame and mindless promoters of superstitions. "You shall not smoke," and that's it.

Who is the "most qualified person" to make certain informed decisions may be fun – and controversial. On Sunday, my non-biological nephew would announce he wants to be hired by the (professional) Czech armed forces, probably to go to the Golan Heights (where our president wants us to fight along with Israel and against the will of the U.N.). It requires some physical and health tests – about 10% of the applicants are admitted. I am afraid that he is much more likely than 10% to be among them. One of the recipes for the blood tests etc. was not to put sugar into his tea for one week. So I was laughing. You know, the sugar in your tea is the same sugar that you're getting from lots of other foods, anyway, and if your body can't reduce the sugar level within hours and all other things within a day, it's probably because you have diabetes and you shouldn't try to mask it. Of course, James confidently said that he would surely be believing experienced soldiers and not me in similar matters. Because of his muscles and mostly because of my diplomatic manners and the anti-science attitudes (selectively focused on my science credentials but not only those) among my relatives, I only replied to myself, silently. "The alternative attitude, James, is to think that a doctor of natural sciences understands the human metabolism more than a professional soldier does." ;-) It's his business, of course, so I think it's ultimately right for him to choose this semi-suicidal path and to reproduce all superstitions that make him a full-fledged member of the squad.

(Some people may be nervous because they see that their scientific knowledge isn't used. It's very bad if they try to inflate its importance just in order for them to prove to themselves that they didn't "overstudy" something.)

Despite the "quantitative" differences in the certainty and many other differences, the case of the climate alarm is indeed analogous to the conversations about health. The society – including people responsible for power plants and many other things – simply have to ask about the costs and benefits of the hypothetical statistical effect. They must demand numbers analogous to those "11 minutes" above. And just like in the case of smoking, they are not getting any of these numbers. Of course that the climate alarmists aren't specific about these issue. Whatever explicit they would say about these matters would further weaken the case for the climate alarm. The expected losses are small relatively to the costs of the mitigation and even if these losses were positive and detectable, they would almost certainly be smaller than the benefits of a CO2-richer (and slightly warmer) world.

So the doctors, climate scientists, and general advocates of the scientific method applied to complex enough questions are right that the science matters even though the effects aren't black-and-white but just statistical. They are right that science may be relevant even though its opinions about a question are always uncertain to some extent – they may be almost certain in some situations and very uncertain in others. They are right when they disagree with all the naive claims of the kind "unless the scientific proof is 100% rock solid and accurate and/or the claimed causal mechanism is 100% certain to work without exceptions, we should ignore the science".

But what is more important is the patient's – and the power plant director's – insight that even though science may in principle relevant in all these complex situations, the recommendation based on cherry-picking a particular effect and identifying it as the ultimate hell – without any efforts to see it in the context, describe the effects quantitatively, and compare them with other effects – is likely to be inappropriate for the real-world decision making process they are responsible for. Just because someone is a physician or has a degree doesn't make his mindless repetition of oversimplified slogans intelligent.

So I agree that the analogy between the climate alarm and doctors' recommendations is conceptually justifiable, despite the much higher uncertainty about the CO2-related climate questions. I just happen to disagree with the mindless adoption of the simplified expert-sponsored slogans even in the case of the doctors.


  1. Lumpencafé !!! ;-)

  2. Politically connected Michael Mann is the climate si e bullet theory reincarnation of Ancel Keys who singularly corrupted nutrition and physiology continuing to this day, decades later, a near perfect analogy to how corruption requires mere careerist enablers instead of a vast conspiracy. Like the carbon footprint counter, the Food Pyramid demonized paleolithic diets, and helped lead to an obesity epidemic from processed carbohydrate loading.

  3. The decision to smoke, in spite of a Doctor's advice, is a personal one affecting one party, the person making the choice. They weighed the subjective utility they get from a longer expected life span against the subjective utility they get from smoking, and found that the latter was greater than the former. The decision of the smoker as long as they are well informed, is always the correct one, in the only meaningful sense that matters. This is why it would be incorrect to empower the physician to override the decision of the patient. People who make this analogy, implicitly believe this not to be the case: they believe Doctors should be able to make armed thugs called the Government stop their patient from smoking by force. Creepy, huh?

    The analogy to Climate Change/Global Warming, is quite strained. In theory, Climate Change represents the effects of the decision making parties on various third parties who have no say in the decision that affects them-an "externality." In this case, the use of force to stop a person making decisions that might harm others who have no say in the decisions, might be justified. The issue is whether one's decisions do in fact harm third parties, and if so, to what extent? The answer appears to be "No, none at all." In which case, force is not justified here, either.

  4. From 20 to the age of 35 I was a smoker, a pack a day and more. My long process of stopping showed me that the "pleasure" of smoking depended on several aspects.

    1) satisfied ritual behaviors ( people have a lot of rituals related to grooming as an example, rituals calm because no decisions are to be made)
    2) "I am as independent as anybody" , emancipation against parental authority too.
    3) an alertness induced by coffee and nicotine in smoke

    4)making smoke rings :)

    5)After I stopped smoking I used to light a match in the morning at my desk, and watch the flame , and that showed me that also there was the issue of "lighting a fire" , surely a survival trait.

    The reason I stopped smoking was because I loved swimming, and I discovered that my stamina was very low. I would easily get out of breath and be dizzy after the swim. I scientifically realized that this was worse after a cigarette :).
    That was my incentive. No doctors no articles or outside input. My love for swimming was stronger than my addiction to nicotine and the cigarette.

    I agree with Werdna that the decision to smoke or not is on the individual, though they are making it harder and harder for smokers, in the winter they have to choose between pneumonia and a fine ! :). Global warming taxes and decisions on types of energy are more or less dictatorial for the individual. In the EU it is not even a matter of voting for the correct party.

  5. coincidentally at the supermarket this morning two men were discussing smoking and stiopping smoking. One of the two said : I want to die while smoking !

  6. "You know, the sugar in your tea is the same sugar that you're getting from lots of other foods, anyway"

    The sugar might be the same, but it's generally accepted that liquid calories are problematic. Our bodies are not very good at registering them and you can easily increase your caloric intake without feeling full just by increasing the percentage of liquid calories.

    That being said I agree that when you're not able to decrease your blood glucose level to normal after several hours you may have blood glucose problems and skipping sugar in your tea (probably) won't help.

  7. Thanks for the point. Maybe it was just about the reduction of the calories - although I am not sure why he would need it before the tests.

    An alternative explanation was that he may have taken some drugs and this restriction of the sugar is a way to eliminate THC etc. more quickly (because the body is forced to burn the reserves if you don't supply sugar) - but I am afraid that he would need to avoid all sugar in that case, not just the liquid one.

  8. Well except Ancel Keys was a real scientist and did not single-handedly corrupt anything.

  9. In addition to clear evidence for the risks of smoking and high level of uncertainty about CC, there are several other important differences between smoking and AGW. There are significant potential benefits from a warmer wetter climate and more CO2 but effectively none from smoking. The use of fossil fuels also provides immense benefits while any benefits from using tobacco are at best trivial. In addition there is a high cost in not using fossil fuels while not using tobacco results in considerable savings.

    In terms of any risk/benefit assessment the two matters are far from comparable.