Feynman Algorithm

Step zero of the Feynman algorithm is always missed:

0. Spent years dismantling every idea or concept and understanding it intimately.

This is almost always done out of sight to maintain the illusion. When people don't see the hard work, they assume you're a genius, not just working hard. I don't think there are any geniuses as such; just people with drive and enough experience and possibly good brain plasticity :)

I can get a stupid high result on an IQ test. That doesn't mean I'm a genius, just that I read up on how they work.

This reminds me of some martial arts philosophy I heard. It was basically that there were three stages to mastery. In the first stage you fight intuitively but know no techniques. In the second stage you practice techniques, and use your conscious mind to choose which of them to use, and how to execute it. It feels forced, and in this stage it is harder to use novel solutions, because you focus on your list of techniques. In the third stage you have internalized all the techniques and can once again work intuitively.

At first glance the Feynman Algorithm is funny, but it also encodes a lot of wisdom about how to go about problem solving.

For example, I've noticed that oftentimes when people get stuck solving a problem it's because they tried to move to step 2 or 3 without first completing step 1.

That page is full of nice quotes. At the end of the day Gell-Mann did not really capture Feynman's spirit and genius because Feynman himself tells all sorts of stories where he uses physical intuition and playful thinking to figure stuff out. Those strategies can be adapted by anyone, not just geniuses.

Read Gleick, who (I think) covers Feynman's approach and other's reaction to it well.

He didn't read the literature as he liked to be surprised. His sister told a story where she made him do so in one instance; preferred to figure things out himself. Was it inefficient? Certainly, but it added to his legend, satisfied his curiousity and often yielded a more-general solution.

Perhaps it was clarity of thought coupled with his (partially?) hidden self-training and toolkit. Perhaps it was a less-mechanical approach to physics, more intuition than perspiration? To me, Einstein falls into this category as well - the ability to ask pertinent questions and not worry about the answer and the number of apple carts being upset.

Certainly there was an ability to focus that was not common.

There's definitely some truth in the Feynman algorithm. If I had to explain how I did some of my research, I'd probably say it similar to that.

While sometimes the manual work to actually show that it works is quite small, other times it may take a month or even longer just to write down the derivation properly.

Ideas that come up after step 2, i.e. "think really hard", are ultimately just that: ideas. You need to ensure that you're not fooling yourself or that you have a case of confirmation bias (skipping steps in a proof because you think it's obvious while it actually isn't). Thus, I'd say step 3, "write down the solution", is at least as important as step 2.