There are numerous attributes that contribute to effective continuous learning and meta-learning, among them:
- where I found something
- how I found it
- who recommended it
- how long ago the information was published
- the context of techniques
- how surprising the information was to me
I contend that this meta-information is actually more valuable than the information itself.
It’s helpful to think about these attributes to get higher signal streams of information. When I find that a particular blog or person has interesting content, I listen much more closely to what they have to say. While I don’t turn critical thinking off, I don’t need to spend as much time considering the source. Links and ideas presented from a trustworthy source tend to be of higher quality. Finding a good source of information makes it much easier to get good information in the future.
Conversely, remembering meta-information allows me to debug and debunk things that I have come in contact with. When I start to disagree with someone whose opinion I previously agreed with, I also think critically about other things that they said or thought. Perhaps there are other views that they held that are also incorrect, and I’m basing my thoughts on this incorrect information. This helps me realize when my mental models need to shift. Everyone has a bias, and I want to make sure that I understand their bias and that is it not harmful to me. If I realize that a much-read software pundit just started selling bug-tracking software, I might start to examine the quality of his articles because of a potential conflict of interest. Similarly, if I understand that the last time I read about something was five years ago and believe my information be out of date, I might preemptively decide to brush up.
Filtering out theories based on recommender is clearly not using the scientific method, but it’s an effective way to get more interesting, useful, and accurate information in a time-efficient manner. Social media proponents purport that search is dead, or at least the type of search commonly thought of today. Tomorrow’s search will be more personalized. I think that this makes sense–crowdsourcing humans are better than computers at the present for separating remarkable things from the merely great and good, and facilitate the spread of useful memes.
Software developers and managers often subscribe to the belief of best practices. After reading Managing the Design Factory by Don Reinertsen, I don’t think best practices exist. To say that something is a best practice and can be applied blindly to any situation ignores the context. Perhaps I’m using a straw man here. But no practice used in the wrong context can be responsibly effective. While I might read that a particular practice was useful for someone, I cannot be considered a responsible practitioner unless I understand their rationale and the constraints of their project and think that the practice will apply to my situation. For example, if I am following various entrepreneurial types and remember a technique for increasing sales, it is crucial to be able to remember who gave the advice because of the differences between bootstrapped and VC-backed startups. Trying to apply advice from one category to the other will likely have deleterious, if not disastrous, effects. At best, the outcome will be positive even though the understanding was flawed–a lucky break. So meta-information is crucial in making effective decisions based on information.
Asking “Why do I think X?” and being able to figure out the answer is invaluable. The answer is possible with meta-information. Thinking about meta-information forms a structure for information to go into for future review and comparison. Forming this structure allows one to navigate through information overload by being able to quickly gather up-to-date information and ensure that this information is reasonably correct. Do they teach this in school? How does one learn this technique?
It’s interesting to talk with people who can trace their mental models. I typically place a higher degree of trust in someone who knows not only what they think, but why they think it, with relevant citations. I can trace their thoughts and compare them to my own structures. Doing this well requires a focus on the context and validity of information.
Writing this post leads me to consider that the best information sources will actively supply context to reduce the overhead for readers. Useful case studies or research papers supply the problem, the specific conditions of the situation, and the attempted solution with metrics to show the effectiveness. The best journalism paints a broad picture and gives not only information, but context. These are the articles that people reference and point others to, because they are timeless. Every forgettable blog or forum post contains mostly the author’s unverifiable and subjective experiences and musings.
The last thought process that I want to discuss I got primarily from Managing the Design Factory: information is valuable for disproving theories. When I come across something that surprises me, I get pretty excited. The surprise could be a new way of linking disparate concepts or just a fact that does not mesh with my worldview. New perspectives could very well be incorrect, but it more likely that the context is different from my normal context, or the approach is different from my normal approach. The reason surprising information excites me is because it means that I am not wasting my time. Information is most valuable when it doesn’t fit cleanly into my current mental models. This is more along the lines of the scientific method. If I only look for confirmation of my own theories, I will never disprove them and will instead become infatuated with them. However, if I actively introspect and search for information that might disprove my own models, I stand to be more correct in the long run.