First allow me to show you my preliminary data.: http://www.reddit.com/r/bestof/comments/obtqe/its_they_live_these_are_reddit_robots_programmed/

Introduction

I think I may have found a novel use for an old technique, which may or may not have implications for rational decision making. I am open to constructive criticism or even deconstructive criticism if you make a sound argument. Ultimately, I would like the experiment to be put to the test. If you have the supplies and know-how to carry it out, then feel free to try it and report your findings.

The Goal:

•  Catalyze the brainstorming process in a way that increases both the number and quality of ideas made.

Methods:

• Find a problem that needs solving. Unifying general relativity and quantum mechanics is a good, but ambitious, example. Some more likely problems that could be solved are: "How might I solve my relationship problems," or "how can I advertise my company's product to its target demographic", or "what are some ideas to make quick money."
• Find 2-3 rationalists who understand the problem well. They don't need to be expert rationalists; the most important part is that they know the difference between rationalization and rationality. In the QM example above, and in most scientific applications of the method, all players should have access to the experimental data.
• Assign two of the three rationalists to the "brainstormers" group (name subject to change), whose primary concern is to make logical connections between the data to form hypotheses.
• Assign the odd-rationalist-out as the Confessor, whose primary concern, like in TWC, is to preserve sanity. It is the Confessor's job to catch the brainstormers when they make a logical leap or use biased reasoning. Some tactics the Confessor might use are the rationalist taboo, the reversal test, and argument from the least convenient world.
• This is a scaled up version of what the brain seems to do. We need the brainstormers and the Confessor to act the part of the Apologist and the Revolutionary, respectively.
• The Confessor - brainstormer dynamic is interesting in its own right, but I believe it can be improved. Now bear with me, because the optional step is for the brainstormers to smoke Cannabis. Not too much, but just enough so that connections between ideas are more quickly apparent to them. Remember, the goal is to have the brainstormers make many connections. They need to output quantity over quality, while the Confessor picks out anything that is quality and gently guides the brainstormers toward more quality ideas. Think of it like r-selected evolution.

• Ideally, we would split twelve rationalists into four groups of three (the alternative is to use the same group repeatedly). Group 1 would be told to just brainstorm the problem. Group 2 would be told to choose one among them to be the Confessor. Group 3 and 4 would be told the same, but their brainstormers would be given either Cannabis or a placebo.
• A placebo can be made by extracting the cannabinoids using ethanol or glycerine. All that should be left after extraction is plant matter, and the tincture can be used later for medicinal or recreational purposes. The placebo Cannabis and active Cannabis will have to be rolled into joints because extraction removes some of the plant's pigmentation. If you have access to a lab, then you might follow  this procedure  for the extraction; otherwise, use  this guide  for doing the extraction at home.
• If you don't want to go to the trouble of making the placebo, then you may skip the control group and only do groups 1 - 3. It would be nice to get some preliminary data, even if skewed slightly by the placebo effect.
• For data collection, the Confessor will note down any idea made by the brainstormers, marking the ones which were discarded. After a given amount of time, enumerate the data and compare the groups. The hypothesized result is that the smoking group will make the greatest quantity of ideas, followed by the non-smoking partitioned group, followed by the normal brainstorming group. It is also hypothesized that the smoking group will make the greatest quality ideas, due to a combination of the highly creative nature of ideas made while high (explained below) and the Confessor's job of immediately discrediting any faulty reasoning.

Some evidence that the Cannabis route might be a good one to pursue (more references to be added):

• There is evidence that Cannabis engages the mind in semantic Hyper-Priming^1,2, meaning that distantly-related concepts are primed quickly after having been exposed to an idea. For instance, a smoker might quickly respond to the word "fish" with "submarine," whereas someone who is sober might respond with "fin." If I understand it correctly, then this doesn't mean the smoker cannot answer "fin," only that the more distantly related concepts are given a higher priority than they normally would. One can see why this might be advantageous for brainstorming, but I suggest to take my - and mindhack.com's - interpretation of the paper with a grain of salt until someone with access can read it in full.
• *Cannabis allows erroneous perspectives to be rapidly dismissed in the light of new evidence. While high, it is easy to put one's pride aside and say "oops". This is made especially easy if the user has cached understanding of the art of rationality. In other words: they will listen to the Confessor.* <-- (I haven't found any literature to support this claim, yet. It seems true in my experience, but it might not be true for everyone. If the brainstormers prove to be too clingy, we could alter the method by changing the Confessor's name to Kiritsugu and having the brainstormers agree to always defer to the Kiritsugu's better judgement. The Kiritsugu will have to take care to examine its own judgement and only discard the truly irrational ideas).

Anecdotal evidence:

• Artists, writers, and even scientists have long used Cannabis and other psychoactive drugs as a tool to make "insights." I'm defining insight as the connection and/or creation of ideas (erroneous or otherwise), possibly due to hyper-priming. The Confessor, in the early pioneers' case, was usually their sober self. As Hemingway wrote, "write drunk; edit sober."
• Less gifted stoners have been doing this for ages  but they - for the most part - are completely undisciplined, believe in dubious pseudoscience, and/or don't have a rational observer to moderate them.
• This is going a bit meta, but the outline to the outline of this idea was made while I was high. It was the first time I smoked since having been introduced to Less Wrong and "The Way", and I was surprised to find that I still had most of my wits about me. Although I would often begin down paths that were just Rationalizations, I usually caught myself. In the instances where I didn't catch myself, and it seemed like a legitimately good insight, I wrote the idea down for future (sober) consideration.
• One of the good, practical, non-meta insights I made that night was a life plan. My plan up until this point had been to finish my undergraduate degree and then immediately go to grad school, relying on my schooling and a bit of luck to maybe hopefully turn into a somewhat-successful scientist somewhere along the road. The problem is that I suffer from quite a bit of procrastination, in part because I don't know exactly what I want to do. I don't have any strong passions or any real motivation. My college career, so far, has been an uphill battle against crippling akrasia.
• Aided by Cannabis, I finally saw the obvious: I need to make an effort to find a passion. My new plan is to get a job as a computer programmer after finishing undergrad, but to continue self-teaching in Biology and other sciences. I've already taken the first step by having Computer Science as my minor, and I can help my resume along right now by getting involved in open source projects. As for self-teaching, that's made easy by open courseware like that found on Khan Academy, MIT, and other places, and I always have the old-fashioned solution of just reading textbooks. After following my interests for a while and learning what things I really, really like to learn about, then I'll go to grad school with an actual PhD thesis in mind and money in the bank.
• I'm attributing these insights (the life plan, some other ideas I'm not mentioning, and even the hypothesis itself) to hyper-priming and later editing, but they might have just been made because I was focused on the problems. Hence the need for an experimentally-controlled test.

Conclusion

• Cannabis allows connections to be made between concepts which normally seem unrelated. This is an experience commonly reported by users, and experimentally verified. Some of these connections will inevitably be false, but others might be true, and a third party - a Confessor - might be able to distinguish truth from falsehood. Whether the Confessor - brainstormer dynamic is any more efficient or productive than a normal brainstorming session is an open question, and the only way to really know is to test the hypothesis.

References

What I'm missing. To be included later:

• References to the benefits and techniques of traditional brainstorming. In lieu of that, for now, here's  this  and  this .
• More references to Cannabis research.