In an excellent Authors@Google talk based on his book The Birth of Plenty — about the increase in GDP growth that started around the Industrial Revolution — William Bernstein mentions what he calls “The Resource Curse”:

If all your wealth comes out of a couple of holes in the ground, the quickest way to become wealthy . . . is control of those holes and access to those holes. It breeds corruption, and it breeds poor government, and it drains the entrepreneurial spirit. The best way to get rich is to have no natural resources at all. Think Singapore, think Japan.

Could the same be true of science? Could access to resources — say, a lot of grant money or expensive equipment — breed corruption and poor government, and drain the entrepreneurial spirit? It isn’t obvious why not. Surely human nature is essentially the same in both places. This may have something to do with Gary Taubes’ complaints about poor nutritional science in Good Calories, Bad Calories.

I rarely mention in this blog my animal learning research, which in recent years has been about exactly this — with animals (rats and pigeons). The results of our experiments are easy to sum up: When animals have access to rich sources of food, it drains the entrepreneurial spirit. In psychology-speak, when the probability of reward for actions is high, there is less variation in what animals do than when the probability of reward is low. I got into this line of research by accident.

An editorial by William Bains, a biotechnologist and entrepreneur, questions the usual drug development process:

Translational Medicine [about going from research to practice] conferences are full of discussions of PET fMRI, gene arrays and proteomics, far beyond the means of the GPs that see 95% of patients, and divorced from the simple clinical observations that resulted in the discovery of drugs as diverse as aspirin and viagra.

Because this is the way that biomedical research (especially drug research) is done, it is assumed that the features of this process are features that have to be part of the biomedical research process. These include:

(i) that only professionals operating in established organizations can have the knowledge to identify new areas of medicines research;

(ii) that biomedical research can only be done using cutting edge technology, which is enormously expensive;

(iii) that only tests on huge numbers of people can validate a new approach.

None of these is true.

There is precedent for other ways of doing things:

The majority of clinical advances in the last 20 years of dermatology have been made by individuals working outside the mainstream of academic research, but possessing a keen observational eye, strong, skeptical analytical skills and constant contact with patients

On Friday I had tea with Greg Niemeyer, a Berkeley art professor whose medium is games. I wanted to “gamify” the task I have been using to measure brain function. It is a letter-counting task: I see 4 letters and respond as fast as possible how many are A, B, C, or D. This takes about 600 msec — I’ve gotten a lot faster. Each session has 4 blocks of 50 trials and lasts a few minutes. From each session I get an average reaction time. I have been doing experiments to measure the effect of flaxseed oil (high in omega-3) on this task.

The task is quick, portable (requires only a laptop) and provides 200 fine-grained measurements (reaction times) per session. Flaxseed oil, I have found, not only produces long-lasting improvement in brain function (lasting weeks) but also a short-lived improvement that starts an hour or two after ingestion and lasts several hours. I developed the letter-counting task to measure the time course of the short-lived improvement. To measure the time course, I do the task every half-hour or so. The task has also turned out to be good for discovering other everyday events, such as exercise, that affect brain function. So far I have data from about 450 sessions.

It hasn’t been hard. It could be more fun. The more fun, the easier the research and the more likely other people will do it. Games are fun. Can I make the task more fun by making it more like a game? I asked Greg what makes games enjoyable. In rough order of importance (most important first), he mentioned four things:

1. The right amount of difficulty. Too easy we get bored; too difficult we get frustrated. Mihaly Csikszentmihalyi has made this point.

2. Lots of feedback.

3. Varying problems to solve.

4. Color and sound.

I will try adding these to the letter-counting task. I made a simple RT task with elements of #2 (feedback) and #4 (color & sound). It was much too easy but I am sure that #2 and #4 made it more pleasant.

A London Times article about medical self-experimentation.

Fuzzy logic, which started with a 1965 paper by Lotfi Zadeh, a professor of computer science, is an advanced form of engineering; self-experimentation is a kind of primitive science. They seem to be at opposite ends of a continuum. As science advances — as knowledge becomes wider and more accurate — it becomes more and more useful, gradually becoming engineering. Fuzzy logic is an especially useful form of engineering.

A few years ago I attended a talk by Zadeh in the Berkeley Physics Department colloquium series. He showed a little movie of a platform moving back and forth to balance three linked poles. It was staggering that this was possible. It is a classic problem in control theory. Here is an example with two linked poles:

Fuzzy logic has proved especially useful in building control systems. An early example was furnace control; one of the first real-life examples was a Japanese subway system. Many consumer electronic products, especially those from Japan, use fuzzy logic. One of my Omron blood pressure meters uses fuzzy logic, says the box. (Omron now uses the term IntelliSense instead.)

When an engineer builds a control system, he doesn’t start from scratch, choosing from among all possibilities. Rather, he tries to embody in a computer program what a person would do. The program embodies a series of rules. Fuzzy logic provided a new and better language for describing those rules. It “bring[s] the reasoning used by computers closer to that used by people,” Zadeh has said. People use “vague” rules: If you are near a corner, slow down. Now it was easy to add such rules to control systems.