Cerrisian Blog

It’s easy!

1. Do background research in an area that interests you
2. Find the fundamental papers, re-implement them
3. Try to think of fundamental things you could change, do research to see if anyones done them
4. If they have, take their papers and re-implement them
5. Find a small way that you can change a fundamental assumption or extend it
6. Implement your extension and compare
7. Write it up and publish it

Now, I just need to explain heursitics and techniques to do each one of these

Today I realized that I was using my favorite time management system autofocus and superfocus [link] incorrectly. I was only going through one page at a time, instead of cycling to the next page when I reached the point of “nothing standing out”. Needless to say, things are much better now! I’m no longer stuck staring at bad tasks on a page.

One trick that I’ve finally absorbed from mind tools is that of “recall” after reading material. In short, after reading a section of material, I go over it in my mind and determine the fundamental concepts. Then I generally record this understanding in my own words in a mind map. This is also the point where it would make sense to put it in a spaced repetition system.

I’ve been thinking of how to do GUI mockups, and I realized that videos are a great way to do this. Thus started my open source animation odyssey! Synfig looks pretty good.

In the same vein, I’m also obsessing about paper prototyping and GUI mockups. Which leads me to denim and pencil for firefox. Basically, they allow for rapid development of hyperlinked mockups.

Now, it’s time to go back through my “at home” autofocus notebook.

Flow is that state of consciousness that you experience when you are “fully in the moment” when all of your attention is focused at the task at hand. Where you feel fully alive and remember everything distinctly. Time passes slowly, yet quickly. This is the state that is associated with “optimal experience” a form of happiness by doing and experiencing.

Flow requires goals, clear feedback, balanced challenge and concentration to arise. Any activity can be turned into a flow activity: just create goals and “rules of the game” that allow for feedback and challenge.

We humans are better at handling a large number of small rewards than one large lump sum. This is the opposite of what is provided by power-law distributed activities. You fail and fail and bleed until you get a tremendous payoff. Creating rules and a game of the process can help somewhat, but only if there is no psychological torment provided by your peers. Thus the importance of having a supporting group of loved ones and friends. The rules used to provide “flow” should provide a large number of small wins, good feedback and challenge that is tweakable over time.

Pleasure is linear. We expect to be productive at activities where more input equals more output. That we are coming closer. Thus the difficulty of dealing with things that are a discovery process, where you don’t know that you’ve succeeded until hundreds or thousands of failures. One way to conquer this is to measure output, to measure trying instead of success. Knowledge is subtractive. But we need a narrative way to explain this to others.

Tunneling, focusing, is “L-mode” type thinking. It’s when you’re able to concentrate on one thing and ignore everything else. Pattern recognition and simplification rule here. You see things as you think they are, not as how they are. This state of mind makes us more vulnerable to black swans.

Defocusing is the opposite of tunneling. Sometimes it looks like undirected trial and error, purposely doing things imprecisely, combining things in unusual ways. When we think this way, we reduce our vulnerability to black swans. It also encourages creative thought, aggressive tinkering and exploitation of positive black swans.

• Using an active learning SVM to classify or learn about PubMed articles
• Data mining for various practical solutions, empirical evidence based on articles
• Try to find negative advice. Things that people shouldn’t do to be healthy
• Look in the long tail. Look at the papers with few citations. Take advantage of the fact that citation level is partially a lottery. Look for the gems no one else is looking at.

We are emotional and feeling creatures. We’re just really good at explaining our behavior after the fact. Happiness doesn’t come from gaining new things, but from being in a state of flow. Following our strengths puts us in a state of flow.

We also have free will, it’s just that our resources for self direction are limited. Our locus of attention can only exist on one thing at a time, even if it’s our own thoughts or nudging us to do something. It’s easier for us to snowball changes or actions than it is for us to do things in a big bang.

When bad things happen, we feel better when we know they’re inevitable.

Our outlook and expectations alter our results.

A wiki for handling all the presentation and supporting material for a course, allowing for online editing and correction, diagrams, illustrations, equations and annotations.

Slides and prose. Comments, questions.

It would be fun to use WordNet and language parsing to parse natural language commands in an adventure game.

So, I’m thinking of something like this:

Command -> [chart parser/link grammar] -> [predicate calculus] -> command intent

Use something like WordNet disambiguation to map words to known concepts/words

word -> spelling correction -> wordnet synset -> concept

So, the lucky use this heuristic: When something bad happens they imagine something worse occurring and feel relieved and luck that the bad thing didn’t happen.

Eventually I want to integrate this heuristic into my thinking.

Taking notes, copy and pasting snippets of code that you’re following the thread of into a notepad is helpful. Especially useful for understanding large pieces of code, it allows you to maintain context across multiple methods and classes.

So, the thing that’s running in the back of my mind is that I’m going to use my achiever talent theme in combination with my learning themes to just do stuff. Keep track what I’m learning, incrementally do things, maximize what I’m good at.

In that vein, I’m going to slightly alter my 30/30 work cycle. The 30 minutes of focused work does very well, but I need a better off-beat thing to do. Spending just a few minutes with my eyes closed meditating seems to be really effective, as does socializing or just writing notes, focusing on one article helps.

Another piece of information that is implicit in my emotional intelligence book, but was explicit from a review of a book I read. Calmly focusing on what you learned from anger or frustration is more effective than acting out aggressively. So, that’s a heuristic that I’m going to add to my repertoire and see how effective it is. That’s implicitly in the emotional intelligence book as each emotion is there to tell you something, what is it trying to tell you?

The other thing I have to keep in mind is not to try and do too many new things at once. I should only have ONE habit that I’m focusing on at a time. Right now that habit is keeping a sleep log.