Evolution of business: Variation, Part 1
by Richard BarrettThis is part 4 of an ongoing discussion.
From the initial discussion you may remember the three fundamental functions that evolution uses:
- Replication
Variation
- Selection
Previously… we discussed how replication simply means “do today what you did yesterday.” Sounds simple, but in fact it is very hard. See previous blogs for a discussion of how to replicate your business operations.
Today… we start exploring variation – Evolution’s version of product development.
Change
How does evolution create new versions? To understand the function of variation, first we need to investigate change itself.
Evolution is very careful about how it allows changes in the genetic code. Evolution has even developed—evolved—specific processes to mix the genetic code. Presumably, evolution tried other methods to mix genes but those methods simply did not work. So now we can explore the methods which have survived with the presumption that they actually work, and others simply don’t.
Small changes – Most of the time environmental change is gradual. Average temperatures move up or down only a degree or two. Mean sea level rises by only a few inches. Average rainfall changes only gradually. Seasonal changes are often much greater than the long-term trends. The tidal flow, typically a few feet, is much greater than any change in mean sea level. Daily and annual temperature swings are much greater than changes in long-term averages.
Organisms are designed to survive and thrive across large daily and seasonal variations. Therefore, changes in long-term averages do not require large-scale changes in organisms. Perhaps some tweaking will suffice. For instance, as the mean temperature rises in some high-altitude regions, the first response of many species is simply to move a little higher, where the temperature is cooler. Over time these species will adapt to thinner air, different rainfall patterns and other localized differences in the environment at a higher elevation. All these changes are small, incremental adjustments.
To adapt to these small changes, evolution encourages random changes in individual proteins in the DNA string. DNA uses four proteins (A,C,G,T) in its genetic code. Evolution uses everyting from x-rays to intentional transcription mistakes to create single-protein changes. Think of these changes as single-letter changes in a book. Within a single word or sentence, the change may be large. Consider the sentences below:
Evolution operates over long time scales.
Evolution operates over long time swales.
At the word level, both text strings contain legitimate words. At the sentence level, you probably guessed the intent of the second sentence, even with the misspelling. Within a paragraph and certainly with a page, the change is insignificant. Most single-protein changes are like this—no long-term change in survival and only a short-term inconvenience to the organism.
Now consider the sentences below.
Let’s eat, mother.
Let’s eat mother.
A small change in a single character here makes a big change in the meaning.
Evolution likes these changes. Most of them are detrimental to the organism, reducing its survival. But a few may create a competitive advantage in survival. Evolution is happy either way. Selection removes the detrimental changes in one or two generations while the good changes eventually propagate through the species, improving its survival, even if only marginally.
Large changes – Evolution has developed mechanisms to cut and swap DNA in large chunks. If we think of DNA as computer code or a long string of text, simply cutting and swapping large chunks of DNA code will lead to confusion and disaster. Imagine reading A Tale of Two Cities, and suddenly finding a big section of Mark Twain inserted into the text. The sentence crossing the cut point will not make any sense, and the flow of the story across the cut point will certainly be peculiar.
However, evolution has co-evolved this adaptation and the response of DNA to it. Since it has proven successful in creating new variations, evolution continued to use this process. In response, DNA has generated large areas of blank space (junk DNA) between active sections of DNA. Some DNA analysts estimate that the blank space may be as much as 30-70% of the total DNA. As a result, a random cut/switch of DNA code has a good chance of cutting and switching in the blank space between active sections. Thus this process provides good randomized mixing, with only infrequent damage to active DNA sequences and organisms.
Of course, when the random cut and swap does destroy an active DNA sequence, the consequences may be fatal to that particular organism, but evolution runs millions of tests so a few failures are to be expected, just like product versions that don’t survive in the market.
The Topographic Map of Survival
If we think of a topographic map, we can better understand why evolution has developed two different strategies to create variations, also called mutations.
To climb a mountain, we can usually follow the simple strategy of making sure each step goes uphill. This corresponds to random changes of individual bits in the DNA sequence. Each individual change represents a single step– uphill or downhill we don’t know until natural selection evaluates the value of that step, but each single-bit change is only a single step. This type of change results in a slow, steady path up the hill, one step at a time.
Evolution views the environment as a topographic map, with hilltops representing areas of better survival for each species. Each species climbs the hill by becoming better adapted to its specific environment.
Mathematically, this is an optimization problem. Given some criteria of “goodness,” a species climbs the hill as it moves closer to the criteria of goodness. For evolution, the criteria is very simple – survival. While a complex set of inter-related factors all contribute to survival, the final exam is a simple pass/fail test. Did the organism survive to reproduce? Did it pass on its genetic code? Species at the top of the hill are very well adapted to their environment, so the species tends to last a long time with few changes.
Now go back to that topographic map. You are near the top of one hill, but there is a taller, larger hill nearby. How do you get there? The simple strategy of “just go uphill every step” won’t take you back downhill then uphill once again to the top of the larger, higher mountain. You need a different strategy to jump from one hilltop to a nearby hilltop. Hence, evolutions other strategy – cut and swap big sections of DNA. This approach allows evolution to jump from one hill to another. The combination allows evolution to climb individual hills and simultaneously to jump to other hills to explore the evolutionary landscape more thoroughly.
Next week we will explore the beauty of failures and the two biggest mistakes managers make in nurturing failures. See you then…
