SCALABILTY AND 2 TYPES OF UNCERTAINTY 1. Mediocristan “When your sample is large, no single instance will significantly change the aggregate or the total.” “The largest observation will remain impressive, but eventually insignificant to the sum.” Eg “…round up a thousand people randomly selected from the general population and have them stand next to one another…imagine the heaviest person you can think of and add him to that sample…he will rarely represent more than a very small fraction of the weight of the entire population…” 2. Extremistan “…inequalities are such that one single observation can disproportionately impact the aggregate, or the total.” Eg base the line up example above by wealth rather than weight…one person may end up representing 99% of the wealth Mediocristan is more about the physical, whereas Extremistan is more social quantities (informational)…it can produce black swans (a few occurrences have big impact) Wars are now Extremistan as one button press can kill lots of people, rather than one at a time KNOWLEDGE Mediocristan - more predictable based on past data, any future surprises will not be consequential…more data you get the more you know…tyranny of collective, routine, predicted Extremistan - always be suspicious of knowledge derived from data, as one new piece of data can affect the total massively…tyranny of singular, accidental, unseen, unpredicted…vulnerable to black swans
bit.ly (Short URL)
