A shelf for thinking about cheap intelligence
Books, essays and papers, most of them written long before modern AI, that turn out to be about it.
reading pile. growing.
trying to answer →What is a moat in an AI world?Where does value move when intelligence becomes commoditized?What becomes scarce when intelligence becomes cheap?Why might human-made things become more valuable?
Most of the useful thinking about what happens when intelligence gets cheap was done before it did. It was done about printing, copying, software, search and photography. This is the working shelf, in order of publication, with a line on why each one is here. It grows from the Reading & Watching log.
- 1935
The Work of Art in the Age of Mechanical ReproductionWalter Benjamin
aura: the thing a copy lacks
- 1966
The Tacit DimensionMichael Polanyi
“we can know more than we can tell.” the part not in any training set
- 1966
Performing Arts: The Economic DilemmaWilliam J. Baumol & William G. Bowen
why the string quartet keeps getting relatively more expensive
- 1997
The Innovator’s DilemmaClayton M. Christensen
good-enough from below, except now good-enough is a model
- 1998
Information RulesCarl Shapiro & Hal R. Varian
pricing things that cost nothing to copy, the last time this happened
- 2002
Technological Revolutions and Financial CapitalCarlota Perez
installation, frenzy, turning point, deployment. where are we?
- 2002
Strategy Letter VJoel Spolsky
“commoditize your complements.” value moves next door
- 2015
Aggregation TheoryBen Thompson
when distribution is free, winners own demand
- 2016
7 PowersHamilton Helmer
a vocabulary for moats. which of the seven survive?
- 2017
Attention Is All You NeedVaswani et al.
eleven pages that rearranged an industry
- 2019
The Bitter LessonRich Sutton
general methods plus compute beat cleverness. uncomfortable, mostly right