Artificial Accuracy

Coming from a family of educators, I have become deeply concerned about the state of education in the United States today. My largest issue is how education does not relate enough to innovation, as the only hope for humanity to solve its problems has always been through education. This has been true since we learned to learn, and will continue until or unless we forget how to learn.

The concept of signifiant digits, also called signifiant figures is relevant here. My embedded Apple dictionary defines this as “each of the digits of a number that are used to express it to the required degree of accuracy, starting from the first nonzero digit: this text will round numbers to three significant figures.” The basic concept is beyond a specific point, more numbers do not increase accuracy. Ten divided by two is five. There is only one signifiant digit even though 5.00 or 5.0000 are also correct but without increasing accuracy. Infinitely long series of digits are important in science and engineering but less so for normal business people where in general there is only one significant digit.

It is rare for a business plan to predict anything with better than 10% accuracy, which is why 100 page business plans are not more significant or useful beyond 10 pages. As most plans do not survive contact with reality, there is not much point to being overly attached to detailed results derived from hypotheticals. Innovators and startups are both much more likely to succeed when they are flexible enough to change in response to the external world. Until an enterprise gets into a feedback loop with reality, through engaging real customers, suppliers, staff and the other stakeholders, extended speculation is counterproductive.

Much more than a page or three on predicted revenues for a startup is engaging in artificial accuracy.  And from an innovation perspective so is teaching to a test. There are several problems with it. It presumes someone actually knows what should be on the test that is being taught to, and that implies a universe that is more static than it really is.  Also it is the lower primary and secondary grades who are tracking the most to tests, and these individual teachers have been taught to teach, often do not have much domain expertise which comes from experiences in the outside real world. On the other hand higher education is staffed by individuals with significant domain expertise who are often not trained in teaching or in the communication of knowledge.

There is a problem here – the world is accelerating so much that the half life of specialized knowledge is decreasing every year. In other words when we teach students increasing amounts of information,  or even knowledge which is more context rich, its life expectancy decreases every year.  The only hope is teach people how to learn increasingly quickly accompanied by the critical thinking skills to determine what to learn and what to forget.

We desperately desire rigor for repeatability and predictability which brings us to metrology, the science of measurement where the notion of significant digits lives. It also brings us to a more philosophical approach to life as determining what to measure, and how to measure it, is often not straight forward.

One notion accompanying our changing society is the current definition of quality has been changing as well.

The Metrology of Quality, Quantity and Convenience

Convenience and quantity seem to be becoming society’s new definition of quality. Both require less consumer discrimination than quality. Storing 10,000 songs or photos is considerably easier to understand objectively than determining which ones are great. Is driving 1000 miles better than walking ten great ones?

Meaning, usually more dependent upon quality, is usually achieved more through diminishing returns, than either quantity or convenience. As such quality usually requires more resources in the forms of time, money and effort to achieve. It can also be more difficult to qualify than to quantify.

Metrology, the science of measurement, continues to be critical for humanities development. Rigor requires repeatability, for without the ability to measure how can we repeat and progress? It appears, that for much of the world there has been a shift from quality to quantity, perhaps because it is easier to quantify than to qualify?

This presents us with multiple challenges to overcome – what to measure, how to measure it and how much accuracy is real? And all of these border on the philosophical.

People want real not fake rigor, but they also tend to only absorb information that is emotionally relevant.  This raises the issue of context and context management.

Context and Stories  

Data has less value than Information, which has less value than knowledge. And knowledge has less value than wisdom.   What is increasing along this path from data to wisdom at every level, is context.

One reason stories are so successful at transmitting lessons is they place information and knowledge within an experiential context. This permits making information and knowledge emotionally engaging which is necessary for emotional relevancy, the only kind of relevancy.

This makes it difficult to determine exactly which accuracy is not artificial.