Saturday, January 07, 2006
Can you measure capability quantitatively?
I recently saw a speech at CES in Las Vegas with one of the folks from Google, Larry Page. The speech was read from a set of papers. I them remembered a post on Xooglers by Doug who speaks of his interview where Sergey Brin asks about SAT scores and GPA. And the question arose in my mind, can you really predict perfomance by SAT and GPA? SAT, you see, is arguably an indicator of raw intellectual horsepower. Your GPA might be an indicator of discipline and applied intellectual horsepower, although I suppose that might depend of the rigor of the education that comprised the GPA.
My gut feel is that these indicators are strong on average, but frequently wrong when evaluated on pure measures like impact or personal success as defined by wealth. One might ask, if we were to change the criteria from GPA to Net Worth, would the concept work the same?
In any case, I was amazed that Larry was reading the speech from papers. Here is a guy who is promoted to be one of the brilliant founders of couple who cannot commit a short speech to memory? Maybe he needs to take a test. Bill Gates never finished college, and he almost always speaks from memory. Perhaps a core indicator should be good test scores, mediocre educational performance, but an excellent memory.
The whole point of this article is to suggest that the categorization of people into buckets is foolhardy. Our understanding of genetics and the chemistry of humans is at a primitive level. Engineers can define algorithms because the problem set can be understand and bounded. Until we have a meaningful understanding of the human condition, the guys at Google should avoid stereotyping and categorizing people based on simplistic and incomplete criteria.
My gut feel is that these indicators are strong on average, but frequently wrong when evaluated on pure measures like impact or personal success as defined by wealth. One might ask, if we were to change the criteria from GPA to Net Worth, would the concept work the same?
In any case, I was amazed that Larry was reading the speech from papers. Here is a guy who is promoted to be one of the brilliant founders of couple who cannot commit a short speech to memory? Maybe he needs to take a test. Bill Gates never finished college, and he almost always speaks from memory. Perhaps a core indicator should be good test scores, mediocre educational performance, but an excellent memory.
The whole point of this article is to suggest that the categorization of people into buckets is foolhardy. Our understanding of genetics and the chemistry of humans is at a primitive level. Engineers can define algorithms because the problem set can be understand and bounded. Until we have a meaningful understanding of the human condition, the guys at Google should avoid stereotyping and categorizing people based on simplistic and incomplete criteria.
# posted by John Dohm @ 9:50 PM 
