> Bostrom would no doubt go on a multi-country speaking tour, warning captive audiences we had only a few more years of human culture (**maybe he’s right, in a different way?**).
Got the chills when reading the "in a different way?" part. It's as if LLMs amplify what's been wrong with us humans for quite some time. In a way, it might be a wake-up call, a reminder that our obsession with tech kinda let us forget that we need to obsess with our humanity in the first place.
I think that's a great read on it, yes. It seems we can see technology in a dual way, as a solution to some human problem and often as a mirror to what's wrong. Thanks for your comment! I love this insight.
Thank YOU Erik for always providing so much food for thought! Your articles are always top-notch. I’m really happy I found Colligo, it’s a kind of safe haven for me in the crazy times.
Erik, thank you for this simple and brilliant analysis of the mechanics of a myth, and the specific features of the technological myths, well extracted. I will be attending a conference in Prague on 24.4 April - so pls let me know if it is ok, if I reference these points in my panel talk, of course, quoting you.
Consider the irony. A species with thousands of massive hydrogen bombs aimed down it's own throat claims to be able to create super intelligence. Sounds like a "garbage in, garbage out" operation to me.
Yes, that is an irony. Of course making a hydrogen bomb was no small feat! Thermonuclear fusion on Earth amounts to understanding what the Sun is doing 93 million miles away, and reproducing it technologically and of course using the famous conversion E = MC(squared)--that "squared" sure did a lot of work for us! For me, the irony is that we could make hydrogen bombs, but remain obsessed with calling glorified calculators intelligent. Anyway, lots of garbage in, garbage out! Thanks for your comment.
Very though-provoking👍 Several things come to mind:
- Do we not, each of us, only SLOWLY realize what we are, which is always a rough model of how we function - ergo also what our intelligence is - but it is structurally incapable of recognizing itself reflexively - because so much is taking place subconsciously - which we can at best characterize very roughly by it's effects, e.g. sudden insights out of the blue or powerful emotions (only partly explainable by their triggers but never in the exact way they will feel).
- Were we able to transcend our individual self-understanding by means of shared understanding passed on as culture? The answer is pretty obvious - but any attempt to reproduce our intelligence as AI remains short of AGI for this very reason.
- Mythology. Yes, maybe mythology plays the role of and is most influential in- emotionally fulfilling our need for hope - for advancement (in whatever arbitrary sense). But this ignores the depth of knowledge, ALSO about our human nature (including intelligence), that ANCIENT mythology reflects. I go so far as to say that nothing significant was revealed about human nature beyond the Mahabharata epic (the one that contains the Bhagavad Gita) - at best it was formalized and "sciencified" - but only a miniscule portion of the phenomena described there by way of myriad characters and sub-stories - each a case-study - so we would be closer to AGI if AI knew them all and could explain them - which is not possible except with very deep abduction.
Erik, nicely done as always. Humans, life together, require myths of some sort, shared beliefs -- I'm here concerned with their social . And California, and as you well know, Silicon Valley, has pretty much always been a land of myth. So the social/political/critical question is not "when will we actually have AGI?" but "what does it mean to give ourselves these myths?" As I think I've said to you before, you are well situated to confront these questions. Keep up the good work!
> Bostrom would no doubt go on a multi-country speaking tour, warning captive audiences we had only a few more years of human culture (**maybe he’s right, in a different way?**).
Got the chills when reading the "in a different way?" part. It's as if LLMs amplify what's been wrong with us humans for quite some time. In a way, it might be a wake-up call, a reminder that our obsession with tech kinda let us forget that we need to obsess with our humanity in the first place.
Hi Ondřej,
I think that's a great read on it, yes. It seems we can see technology in a dual way, as a solution to some human problem and often as a mirror to what's wrong. Thanks for your comment! I love this insight.
Thank YOU Erik for always providing so much food for thought! Your articles are always top-notch. I’m really happy I found Colligo, it’s a kind of safe haven for me in the crazy times.
Erik, thank you for this simple and brilliant analysis of the mechanics of a myth, and the specific features of the technological myths, well extracted. I will be attending a conference in Prague on 24.4 April - so pls let me know if it is ok, if I reference these points in my panel talk, of course, quoting you.
https://qubitconference.com/
Hi Jana,
Sure it's okay! And congratulations! I'd love to see the talk if it's recorded.
Thank you.
Consider the irony. A species with thousands of massive hydrogen bombs aimed down it's own throat claims to be able to create super intelligence. Sounds like a "garbage in, garbage out" operation to me.
Hi Phil,
Yes, that is an irony. Of course making a hydrogen bomb was no small feat! Thermonuclear fusion on Earth amounts to understanding what the Sun is doing 93 million miles away, and reproducing it technologically and of course using the famous conversion E = MC(squared)--that "squared" sure did a lot of work for us! For me, the irony is that we could make hydrogen bombs, but remain obsessed with calling glorified calculators intelligent. Anyway, lots of garbage in, garbage out! Thanks for your comment.
Very though-provoking👍 Several things come to mind:
- Do we not, each of us, only SLOWLY realize what we are, which is always a rough model of how we function - ergo also what our intelligence is - but it is structurally incapable of recognizing itself reflexively - because so much is taking place subconsciously - which we can at best characterize very roughly by it's effects, e.g. sudden insights out of the blue or powerful emotions (only partly explainable by their triggers but never in the exact way they will feel).
- Were we able to transcend our individual self-understanding by means of shared understanding passed on as culture? The answer is pretty obvious - but any attempt to reproduce our intelligence as AI remains short of AGI for this very reason.
- Mythology. Yes, maybe mythology plays the role of and is most influential in- emotionally fulfilling our need for hope - for advancement (in whatever arbitrary sense). But this ignores the depth of knowledge, ALSO about our human nature (including intelligence), that ANCIENT mythology reflects. I go so far as to say that nothing significant was revealed about human nature beyond the Mahabharata epic (the one that contains the Bhagavad Gita) - at best it was formalized and "sciencified" - but only a miniscule portion of the phenomena described there by way of myriad characters and sub-stories - each a case-study - so we would be closer to AGI if AI knew them all and could explain them - which is not possible except with very deep abduction.
Erik, nicely done as always. Humans, life together, require myths of some sort, shared beliefs -- I'm here concerned with their social . And California, and as you well know, Silicon Valley, has pretty much always been a land of myth. So the social/political/critical question is not "when will we actually have AGI?" but "what does it mean to give ourselves these myths?" As I think I've said to you before, you are well situated to confront these questions. Keep up the good work!