mathbabe

Well, like many I have much enjoyed your ‘Weapons of Math Destruction’, and have been doing my bit in recommending it to friends. It surely deserves to be widely read, and indeed it would be even better if it garners a prize.
So can I ask for some minor modifications in the next edition, which must surely come.
It might have already have been brought to your attention that Kent, in England, is not a city but a county. Not as infamous as Kent State University in Ohio.
‘Rejigger’ – on page 67 of the UK paperback edition. I had to check this, because in UK English ‘jigger’ is to mess something up; we would say ‘rejig’ for ‘rejigger’.
The use of large percentages: It always seems to me that people have difficulty when terms like ‘a four hundred per cent increase’ etc are used, rather than a simple multiple, saying that something has increased five times. It seems better (to me) to reserve percentages for genuinely fractional increases or decreases, like 110%, or 90%.
On page 209 you talk about positive feedback loops. It would be helpful to clarify how the term is used differently depending on context. In systems, positive feedback usually leads to the system becoming unstable and breaking down; negative feedback is usually a stabilising influence. In common speech, positive feedback is interpreted as encouragement; negative feedback as (unpleasant) criticism. The positive loop that you describe seems more akin to a heuristic corrective, which is exactly what you want in an algorithm, and in terms of systems would be ‘negative feedback’. Perhaps that seems overly pedantic…
On page 129 the reference to the chaotic lives experience by children reminded me of the observations in Richard Sennett’s book ‘The Corrosion of Character’, where he similarly notes similar damaging effects upon family life. (He is describing the States.)
The use, or rather misuse, of algorithms brought to mind two things. First the way that Alfred Binet’s attempt to use measures of intelligence as a way of helping children who were struggling became instead a means of classifying and then discounting these children. Second, it reminded me of a short science fiction story: ‘No Fire Burns’ by Avram Davidson, in which a psychologist devises a test for psychopathy, initially understanding that it will be used to weed out unsuitable employees, but then discovering that it will be used to select people with no conscience who will willingly murder:
https://web.archive.org/web/20010616052038/http://www.scifi.com/scifiction/classics/classics_archive/davidson2/davidson21.html

Tests – and algorithms – are tools which can be used for good or ill. I do agree about your demand that there be a code of ethics.

Thank you so much for your book. I hope it makes a difference.

John Henderson [sorry of this gets posted twice…]

LikeLike