On Reading Joseph Conrad’s Nostromo: critical theory, pedagogy and artificial intelligence, or Why Tversky and Kahneman are wrong

Economist Andrei Shleifer writes in his review of Noble-prize winning, behavioural economist Daniel Kahneman’s 2011 book Thinking, Fast and Slow, that it was “a major intellectual event” and lists Kahneman’s work together with that of Amos Tversky as having had an impact “on psychology, but also on such diverse areas of economics as public finance, labor economics, development, and finance.” In this article I argue that a further link can and should be made also to an impact on literary criticism and theory, pedagogy and thinking about (General) Artificial Intelligence (G)AI, for I argue that the most prevalent models of reading both in globalised education (at all levels) and (therefore) also those that determine the design and outputs of (G)AI are underpinned by these psychological discourses. I argue further not only that it is the case that these models of reading coincide with Tversky and Kahneman’s arguments, but also that this is a profound and crucial problem, for two reasons: first, that there are key problems with their arguments in their own right and, second, that the naturalisation of this model of reading on the basis of their kinds of arguments is implicated in extensive wider difficulties not just for education or (G)AI alone, but also for society and politics at large. To demonstrate the implications I analyse both the problems in Tversky and Kahneman’s classic 1974 article “Judgment under Uncertainty: Heuristics and Biases” and discuss some criticism on Joseph Conrad’s 1904 novel Nostromo.