theorem debunks the most important AI myth. AI will not be conscious, says Roger Penrose (YouTube interview).
Categories
theorem debunks the most important AI myth. AI will not be conscious, says Roger Penrose (YouTube interview).
5 replies on “Gödel’s”
I’m glad Roger Penrose is a physicalist. I like his work with Stuart Hameroff on microtubles. Though I agreed with him otherwise, the fundamental idealism of Federico Faggin’s latest book Irreducible put me off reading it. Despite interest in the quantum fields perspective, I felt he should have acknowledged form and structure for these actions at Planck scale, or just above. Existence being called into tangible being by the mind was too much for me, and rendered consciousness somehow flat; an inverse monism. I was more inspired by reading Jaegwon Kim in Physicalism or Something Near Enough. It’s good at least that Faggin also acknowledges that computers cannot be in possession of a self, and lack understanding (transcendence as Penrose puts it), and assimilation of qualia; lacking in all perception. That said, is it controversial to seriously wonder, beyond euphemism, if all humans are conscious? I wonder that sometimes. I would like to acknowledge it in all of ‘us’, and do acknowledge animals as conscious in general, but sometimes it feels as if I’m in a room with David Chalmers’ zombies, as if some people don’t ‘pass the Turing Test of the heart’ as I put it fairly recently (despite that being a poor metaphor). I’ll just watch the rest of the interview. Thank you for sharing this. Always fascinating.
One of the problems with the Christian worldview is that it equates humans. In the real world, only the most emergent Homo sapiens have free will, which are very few. The rest are slaves to parental introjects.
The same can be said with the degrees of consciousness, from the amoeba to the philosopher. The mass is made up of humans that I call Neanderthals, and only the most emergent deserve the title of Sapiens.
What I like about Penrose is that he says that the king (AI) is naked. It is something similar to what Peter Schiff says about Bitcoin. But the masses believe any nonsense that you put in front of them.
When the financial debacle comes, those who believe in cryptocurrencies will suffer; and on the other hand, that nonsense that we saw in the Terminator series won’t happen (it is prolefeed for the proles!).
I notice the ludicrous, blind nature of AI from personal experience trying out Grammarly recently (I thought I’d see if my books needed any adjustments). The program’s AI-driven ability to choose the right grammar and punctuation seems reasonably sound – if still perplexed by some of my correct formatting – but the word choice assistance is pitifully weak, and you know there’s nothing there to grasp the sentences, and to understand the context. In a simple example, I was repeatedly offered false corrections such as ‘excellent’ where I had used ‘great’ (as in ‘huge’), and at the level of my normal writing it didn’t even seem able to construct adequate style rules to match me. I examine it in action and think: the likes of this could never take over. It’s too stupid. There’s no intuition, and no comprehension. Thus it can’t really get better, it can only be programmed to be more intricate. I’m glad you agree over Neanderthals, despite my long-winded synonyms. Yes, and it’s a shame the dissident right fall for Bitcoin so readily. It’s not even as if it’s discreet (which I sense is the lie that appeals to them), considering the entire blockchain system can be openly observed. But you tell them about gold as I did and they laugh at you and shrug it off…
Where you can tell these machines don’t have a mind is when an “AI” program answers the phone for a service that receives many calls. They are programmed to direct you to a certain number of problems, but when yours is not on their list I get exasperated and start yelling at the computer voice: “Get me a real person!”
This happened to me a few days ago when I called FedEx for a package that, due to my move, I gave the wrong address for my family’s school. The program that answered me simply didn’t understand that and only directed me to a real person until I kept yelling “Get me a real person!”
I empathise fully. What you’ve just described is my every interaction with the NHS in this country since they automated the entire service. My partner doesn’t understand the murderous rages I fly into having to deal with Patient Access/Ask my GP AI booking services – an endless procession of browser windows, 2FA confirmations and phone answering machines, again as you say, all missing the vital queries (and never accommodating of my non-smart technology). It happened to me today simply trying to arrange to speak to a GP face to face (denied in the end; phone is the best they’ll manage for me). Hours literally of navigating windows and phone queues, left in the dark in between. I swore to her I wanted to go in there with a baseball bat (etc.) and ‘go postal’ on them. I wasn’t being serious, but I was angry enough to be. This move to AI automation will be their downfall, I believe. As more systems take it up, society will just collapse. Royal Mail is the same, as is my banking. I imagine there are many people each day who get as angry as I got earlier, at being effectively stalled and insulted. This country is already seething. David Betz’ recent civil war predictions still seem relatively on point.