Performance Enhanced Theorems

Among sports fans there is a lot of controversy about the use of performance enhancing drugs by athletes. Of course what exactly an performance enhancing drug is, is often left pretty vague. But most sports fans argue against the use of such drugs because they are, in some form, cheating.
So what do we do, for example, with Erdos. Paul Erdos was, for those who don’t know, one of the twentieth centuries great mathematicians. He was the author, amazingly, of over 1,500 different papers. It was well known, also, that Erdos used amphetamines. There is a famous story that his friends were so worried about his amphetamine use that they bet him that he couldn’t stay off the drugs for a month. Of course Erdos took the bet and successfully stayed off the drugs for the required amount of time. When he went to collect the bet, he reportedly said “You set Mathematics back one month!” (OK, this story is off the top of my head, the details may or may not be correct!) So, in a real sense, Erdos’ productivity was increased in large part by his use of a performance enhancing drug. So, was Erdos cheating? Should we think of his “records” his “theorems” as somehow being “less proved by Erdos” because of his use of amphetamines? Well I certainly would argue against this.
At FOCS this year I was having this conversation, and it came up that perhaps we shouldn’t penalize the result, because the outcome, i.e. the mathematical proofs isn’t really a competitive sport. But what about those who are competing for tenure? Now of course I’m disregarding the legality of the performance enhancing drugs. But disregarding this issue (which may just invalidate the whole argument, but bear with me) should there be drug testing of tenure track professors to make sure they aren’t using amphetamines to increase their productivity?

13 Replies to “Performance Enhanced Theorems”

  1. Erdos’ publication record shows you the problem with using amphetamines to “improve” your research. Namely, even though Erdos was brilliant, many of his 1,500 papers aren’t very good. If drugs don’t actually gain you performance, but only adulterate it or borrow it from the future, then yes, it should give a tenure committee pause.
    On the other hand, drugs that really do make you better are just fine. For example, vitamin B is a performance-enhancing drug: it prevents memory loss. Everyone should take it.

  2. “But disregarding this issue (which may just invalidate the whole argument, but bare with me) should there be drug testing of tenure track professors to make sure they aren’t using amphetamines to increase their productivity?”
    Yes.

  3. The problem with drugs is that theyre bad for you, and people facing
    competitive pressure may sacrifice their long-term health in a zero-sum
    tenure race. Similarly for sports, where its more obviously zero sum
    (humanity isnt advanced by having more home runs per capita per year). By
    that logic, recreational drugs should only be banned in settings where we
    dont think people can overcome peer pressure or addiction.
    Perhaps a question we should ask is whether the social benefit of increased
    researcher productivity outweighs the health costs borne by researchers. I
    think in Erdoss case, the answer is yes, but that educating and funding more
    scientists is usually a better idea.
    On the other hand, perhaps drugs would make more sense in a situation like
    the Manhattan project. In a few decades, things like renewable energy, clean
    water and cures for global warming-induced epidemics all might warrant
    sacrificing the researchers health.

  4. Sorry, I missed the question (which may not have been meant seriously) at the end of Dave’s post. All I meant is that if a tenure committee knows that the candidate is using amphetamines, it raises questions about his or her judgment and likely long-term productivity.
    Pro-active drug testing is not such a good idea.
    Except in the blogosphere, where it shines compared to many other ideas that have been proposed. 🙂

  5. Heh, that story about Erdos is great. If we start drug testing, we should do it to tenure professors too. There were a few I’ve had I’m pretty sure were on crack.

  6. Amphetamines are clearly very dangerous and I would be rather afraid to be around anyone using them.
    However, the drug of choice these days for math/CS types seems to be provigil (modafinil) which seems to not have some of the unpleasant side effects of amphetamines and is not addictive. Is that an unfair advantage? Probably. Is it unhealthy? Seems to affect the liver. But great when you have a deadline!!

  7. There are several ways of messing up your brain without resorting to substance abuse. Two unslept nights in a row, twice a week, over three months, should be enough to make a genius of clarity of vision out of any idiot. Drugs may make your work feel brilliant to yourself, but I am not convinced that they actually enhance mental performances. A connected question that I find interesting is: what is the link between intelligence and mental illness? Is there any computable trade-off, and does, uhm, Zen for instance relax it? Is brain a self-regulating system? Ultimately, are, say, words, constructs or entities? Do we emit them or do they modulate us? And why did the egg cross the street?
    Nice entry, Dave.

  8. Caffeine, nicotine and chocolate are by far the most widely-used stimulants in Western society. They are so deeply ingrained in our cultural mindset, our history and our economies, however, that their comparable usage is left unaddressed in most discussions on this topic.
    Do we question the ethics of a morning cup of coffee to sharpen one’s alertness? Attention focuses upon unfamiliar compounds due to novelty and the absence of socially-defined norms surrounding their usage. Any novel compound which modifies the dynamics of the nervous system raises broader philosophical questions – and accompanying insecurities – due to its consequences for our definitions of mind and self-identity.
    With the accelerating pace of bioscience, the ethics of using these ‘cosmetic’ compounds should be addressed now – because today’s newest discoveries will be seen as primitive and crude analogues in comparison to future compounds that will be more highly specific, minimized in side effects, and radically more effective in modifying targeted behavioral attributes.

  9. I don’t have a comment, just a question.
    Considering that amphetamines were widely used during the WWII era and that there was intense pressure for results during the Manhattan Project, did Manhattan Project physicists use amphetamines routinely?

  10. “Amphetamines are clearly very dangerous and I would be rather afraid to be around anyone using them.”
    You’re being rather dramatic I would say. Haven’t you ever heard of Adderall? I know that the prescription of stimulants to kids is particular is another discussion entirely, but whether you like it or not, it is actually a fairly common occurance for children to take amphetamines (Adderall) on a daily basis over a span of years. And these children aren’t dangerous at all.

  11. I cannot see your site properly on my iPhone (I have the 3G, not 3GS). Anyway, I have put your RSS into my laptop, so thanks!

Leave a Reply

Your email address will not be published. Required fields are marked *