Pinker said he had argued that any intelligent system is a mechanism good at solving some problems but not others, and that it is inherently limited by knowledge about the world attainable only through observation and experimentation.

True: the text of Enlightenment Now (2018) contains an argument, made by Pinker, that any intelligent system is a mechanism that is good at solving some problems but not others, and that it is limited by knowledge of the world that can only be gained through observation and experimentation (e.g. in its discussion of AI risk). Both parts must be present in that book. Mostly true: both ideas appear in Enlightenment Now, with slightly different wording or emphasis, or one part is only implied. Mixed: only one of the two parts appears in Enlightenment Now, or both parts appear but in a different Pinker work and not in Enlightenment Now. Mostly false: Enlightenment Now discusses AI but makes neither argument in substance, which appears only elsewhere. False: Enlightenment Now contains nothing resembling this argument, or argues the opposite.

Set before the research began.

1. E1 We're told to fear robots. But why do we think they'll turn on us? Popular Science · 14 February 2018 Own record

   “Computers may be programmed to take on some of these problems (like recognizing faces), not to bother with others (like charming mates), and to take on still other problems that humans can’t solve (like simulating the climate or sorting millions of accounting records).”

2. E2 We're told to fear robots. But why do we think they'll turn on us? Popular Science · 14 February 2018 Own record

   “Understanding does not obey Moore’s Law: Knowledge is acquired by formulating explanations and testing them against reality, not by running an algorithm faster and faster.”

3. E3 Pinker on progress Mindblog (Deric Bownds) · 6 March 2018

   “Chapter 19 deals with existential threats such as nuclear war, the imagined dangers of artificial intelligence, etc.”

4. E4 Steve Pinker on AI Doom [My take: a Rube Goldberg confabulation of Brobdingnagian proportions] New Savanna (blog) · 20 March 2023 Own record

   “This is in contrast to what intelligence really is: a gadget that can compute particular outputs that are useful in particular worlds.”

The Popular Science excerpt from Enlightenment Now (E1) says computers may take on some problems, not bother with others, and handle still others humans cannot. This matches the 'good at some problems but not others' idea, though it is framed around computers rather than 'any intelligent system.' The same excerpt (E2) says knowledge comes from formulating explanations and testing them against reality, not from faster algorithms. That is close in substance to the limit set by observation and experimentation, but worded differently. E3 independently notes that the book's Chapter 19 deals with existential threats including the imagined dangers of artificial intelligence; the two ideas themselves appear, with somewhat different wording and emphasis, in the 2018 Popular Science piece linked to Enlightenment Now [E1][E2].