By contrast, if success was driven disproportionately by a few early downloads, subsequently amplified by social influence, the outcomes would be largely random and only become more unequal as the social feedback became stronger.
So if you have 10 hipsters out of 30,000 as opposed to 9, there is a better chance that the songs that make it up to the top of the charts are wildly different. This does not constitute randomness. Having more people who appreciate one sound as opposed to another will give you a different result, and quite predictable. Unless of course the randomness is with the number of hipsters you are likely to get per sampling of population.
Then this PhD Cornell Microsoft person should not be throwing around the word 'random' in a highly publicized periodical when it is clearly not the term that suits the situation. Alternative terms would include seemingly random, deterministically chaotic etc... but certainly not random.
Comments
By contrast, if success was driven disproportionately by a few early downloads, subsequently amplified by social influence, the outcomes would be largely random and only become more unequal as the social feedback became stronger.
So if you have 10 hipsters out of 30,000 as opposed to 9, there is a better chance that the songs that make it up to the top of the charts are wildly different. This does not constitute randomness. Having more people who appreciate one sound as opposed to another will give you a different result, and quite predictable. Unless of course the randomness is with the number of hipsters you are likely to get per sampling of population.
It's an avalanche effect. The outcome may not be random as you point out, but it certainly looks so.
Then this PhD Cornell Microsoft person should not be throwing around the word 'random' in a highly publicized periodical when it is clearly not the term that suits the situation. Alternative terms would include seemingly random, deterministically chaotic etc... but certainly not random.