Iain McGilchrist gave a keynote speech to the “AI (Artificial Intelligence) World Summit 2022” in the full plenary session on October 12th in Amsterdam.
The Video is available here and the full transcript here for members of Channel McGilchrist. [Can’t embed the Vimeo here.]
I thought it significant that Iain chose to redefine “AI” for the purposes of his talk as Artificial Information Processing – presumably to distinguish it from any natural or living forms of information processing? I was moved to add this comment:
The summary of his work and his position on what we can do to encourage more right/left balanced approach to the world – and why – is good to see and already well known and agreed amongst those of us familiar with his work.
In terms of the relation to “AI” and the prospect of AI-enhanced human “Cyborgs” it is telling that Iain has his own preferred translation of AI – “Artificial Information-Processing” as opposed to the “Artificial Intelligence” of the conference organisers and most participants. I agree and consider it very important that Iain follows-up on this:
Artificial Information Processing (Computation) using human-devised machines – qualified as “Artificial” correctly implies the opposite, a “Natural” form of computation too, information processes that happen naturally in living things and natural systems as “Natural Computation”. Iain’s sense of the sacred has led us to Natural Theology or Sacred Naturalism already. Even in the abstract, many information and computing scientists talk in terms of “Machines” – Turing and others – which suggests, even misdirects thinking to, artificially-devised physical machines, but in fact there is a growing body of work that properly recognises information processing / computation processes of systems generally including natural living systems, brains, hemispheres and their neuro-sub-systems. I believe this work already supports Iain’s work on different hemispheric behaviours with not just evidence of the facts and of the mental & behavioural consequences, but with descriptions and explanations of the internal processes by which different-thinking and different world-views arise in the one bi-cameral brain. (I’m thinking “Active Inference” of purposeful living systems after Friston and Solms for example.)
I was intrigued enough to watch the video (generally I avoid them). It’s odd that McGilchrist just drops the distinction as a high-handed correction, with no explanation. My guess is that he wants to leave room for a definition of “intelligence” as more than “information processing.” But unless he’s explained himself elsewhere, we can only speculate.
Yes, that’s exactly why I believe he does it – plenty of earlier evidence. But my point is – the point he’s missing – by qualifying it as “artificial” he is denying a view that even natural processes “process information”. Friston and Solms work would support his, if he let it. It’s significant because others with aversion to the “computation” perspective, like Damasio, have eventually come around to appreciating it.
(Dogma. Information processing is artificial. Intelligence is natural.)
I see it differently. He might allow that information processing is something that both mechanical processes and natural processes can do, but maintain that in either case, intelligence is more than mere information processing. All intelligent behaviour may be based on information processing, but not all information processing is “intelligent,” whether we mean that in a formal or a vernacular sense.
Well yes – obviously – “it” is more than “mere” information processing and not all information processing is intelligent – no-one is suggesting otherwise – all of which can be said without natural or artificial qualification. (I just did).
But why should that make it hard to see that intelligence – and all the different levels and layers of mind and brain sub-systems – does in fact process information?
He’s clearly making a point with the “artificial” qualifier – quite independent of the above.
Well, he is talking to a forum on _artificial_ intelligence, so of course he uses the word. I don’t think he’s using it to make a point. . His emphasis is clearly on the noun. He seems to suggest, with awkward diplomacy, that the forum thinks it’s been talking about artificial _intelligence_ when really it’s been talking about artificial _information processing_.
This doesn’t negate your point that intelligence, whether artificial or natural, processes information. As far as I noticed, the subject of natural intelligence and information processing doesn’t come up. McGilchrist merely draws a distinction between “intelligence” and “information processing” in the forum’s context of artificial instances.
As to the nature of the difference, he doesn’t appear to elaborate, but his mind is subtle. He uses the lecture to extol a better left/right balance in our approach. Perhaps he’s hinting that the forum’s model of “artificial intelligence” over-specifies the role of computing, or artificial _information processing_, in true intelligence.
Ha, we’ve got to this point before. so yes,
“Obviously” “he’s suggesting forum thinks it’s been talking about artificial _intelligence_
when really it’s been talking about artificial _information processing”
“Obviously” “the subject of natural intelligence and information processing doesn’t come up”
“Obviously” “he’s (more than) hinting the forum’s model of “artificial intelligence” over-specifies the role of computing, or
artificial_information processing_, in true intelligence.
Clearly, I’m suggesting his choice of words suggest a point he’s not actually / intending to make. Why otherwise would I post 😉
The contrast between “true” (ie natural) intelligence and the topic of the forum.
He misses an opportunity to contrast left and right brain types of processing.
Sorry, I certainly didn’t mean to belabour the obvious. When you suggested he made the remark about artificial information processing “presumably to distinguish it from any natural or living forms of information processing?”, I must have misunderstood. Obviously it’s “a point he’s not actually / intending to make.”
For what it’s worth, I don’t think the question of natural vs. artificial was on his mind, and I don’t think anything he said in the lecture commits him to a distinction along those lines. Not only did he not actually make the point, he presumably _didn’t_ “distinguish artificial information processing from any natural or living forms of information processing.”