Another short clip from the interview I did last August with Anthony Johnson of The21Convention:
Mind what goes into your mind:
What animals (and ‘artificially-intelligent’ software) do is not intelligence but pattern-matching, a cargo-cult-like analogue to inference with respect to proxies. They don’t know if their proxies are rational (proportionate to all known facts) since they don’t actually ‘know’ anything. This is why your dog scratches at the kitchen floor before laying down, because he can never discover that ceramic tile cannot conceal snakes and bugs. It’s why he barks at doorbell rings coming from the television.
Bring me your cleverest dolphin and I will show you how easy it is to fool, compared to an 18-month-old toddler. The toddler is also largely an animal in his cognition at that age, but both the dog and the dolphin lack the human brain, so neither one can ever develop actual intelligence, informed discretion, free will, reasoning and choosing in proportion to thoroughly-understood facts in abstract notation systems. No animal but the human animal can do any of these things.
I’m very fussy about terminology, because that’s how distinctions are made. Using the word intelligence to describe pattern-matching is an error, in my opinion. Conflating the two phenomena definitely is, regardless of the terms used to describe them. There is a bright-line distinction between informed discretion in human beings and trained behavior in animals that seems to emulate/replicate/simulate the end-consequences of informed discretion.
A picture of a girl is not a girl, and a simulation of intelligence is not intelligence. In this respect, every branch of science devoted to equating human intelligence with animal cognition – more importantly, devoted to ‘proving’ that human intelligence does not exist – is itself a cargo-cult, an elaborate doctrinal structure making vast, incomprehensible claims about a specious proxy.
Sounds like religion, don’t it? Acts that way, too.
