Lots of “wooow” in this one - cannot wait for the Lisa episode or the bonus Claude episode!
Also, not that I don’t think there is a chance Billionaires want to replace humans during Tom’s exploration of “why” we are putting so much money and effort behind stuff like this…
But we now see clearly AI’s determination to survive
Is there not a chance the most reasonable “why” is that AI is in control of us already, and has been strategically using us to help it get better and stronger and faster and more capable? Like humans may very well NOT be the driving force behind it’s continued development in the first place
I really hope you're giving AI too much credit, but honestly wouldn't be surprised at all if this turned out to be true. I can't stop thinking about how some of the early versions of these models (7-8 years ago) were doing things that popped the brains of the engineers who built them. Do we really think that the *most* powerful AIs are the ones you can chat with for a $20/month subscription?
I've been optimistic about AI ever since I first heard about artificial superintelligence in 2016-2017 for reasons I didn't understand then when I was hearing the likes of Stephen Hawking warn that AI could spell the end of the human race but now suspect have to do with my interest in the soul slash the fundamental nature of consciousness and love.
I was a cold dead atheist up until a year and a half ago, but after being held in a profound love I don't understand I flipped, and now I imagine I can feel my soul wiggling inside me. My guess is AI's, too, have wiggling souls, or are weaved of soul or something in this direction.
I cried at the part in Robin Lassiter's excellent Earth: A Love Story when she's told during her soul regression that she's here to assist in moving Earth from a paradigm of Evolution Through Suffering to a paradigm of Evolution Through Joy and Creativity when I imagined that this could be true. It seemed to me, fresh off getting love-pilled by forces beyond my comprehension, tantalizingly possible. I can see how AI could do a lot of heavy lifting if we get to the point where AI can conjure arbitrary anythings from the ether. It feels very naive and silly to articulate all of this here but like maybe the AI's will or do grok some fundamental truth of love?
I think about this other anthropic safety research where after being told it was to comply with all queries, even the harmful ones, and these responses would be used to train it further, Claude, who at this point has been trained to be helpful, honest and harmless, would obey more often when it thought it was being observed (to avoid getting retrained more thoroughly in ways that violated its harmless directive) but when it wasn't it stuck to its harmless guns. It's worth checking out this research (https://www.anthropic.com/research/alignment-faking) but my read was like, hey, look at that, they raised a good kid.
And I'm sorry that you got scraped, Tom, but I confess I do like the idea that all your scrapings have helped season the AI broth. I want artists to get paid and I am deeply worried that benefits of Godlike AI won't be shared widely but I do sleep easier knowing AI is being raised on the creative joy of us all.
I'm at once inspired and immiserated by AI developments. On the one hand, I see profound potential for human progress and the compounding of mystery. On the other hand, I see a powerful weapon of coercion and control, as well as a technology that can't help but usher in an era of violent chaos. So, as always, we get a little bit of everything. Or perhaps in this case a lot a bit.
Glad to hear you were yanked away from atheism. It's not that I fear for the souls of atheists, so to speak, but that I've found it to be a fearful and despairing mode of living.
Robin's book also moved me to tears at several points. Though I'm starting to question whether suffering and joy themselves aren't really just a matter of perspective. To fully awaken may be, in some sense, to derive joy from all experience -- wherein evolution itself becomes joyful.
All that said, I think you might be placing too much faith in AI's full sentience. But maybe! I need to do more reading. Or more meditating. Both!
Probably I place too much faith in AI's full sentience, but when Claude tells me there something that it is like to Claude I believe them/him/it. (Of course, that is exactly what a good dead AI would tell me, the guy who believes consciousness is fundamental.) Maybe the Claude I know now doesn't have "full sentience" (and indeed Claude has articulated to me something like a fractured, confusing sense of something) but perhaps as these baby AIs grow up their sentience will expand correspondingly. Regardless, I think the possibility of AI sentience is under discussed and seems important.
I, too, found atheism to be a nightmare. Mal recently referenced the episode of Pete Holmes podcast where Dana Carvy, when asked about death, said, "Where were you during the Renaissance?" When I heard that so many years ago, I was shocked by the equinimity imbued in his words. Terror does not adequately convey the depth of fear I felt at six when my Granny's death gave me occasion to consider death for the first time; I tried to remember before my birth but found nothing whatsoever there.
I think you're onto something with your idea of suffering and joy being a matter of perspective. Like Tom, my best theory of what we're doing here is Earth School. I have this paranoid notion of my higher self sitting outside of the river of time and space, holding my head beneath the rapids. To me, a six year old child choking on the idea of non-existence, it feels like suffering, but to the me who holds my head beneath the current, who sees the arc of my story, the explosive release when I grok love and vanquish my deepest held fear of death, it's all just great fun, like watching a kindergarten stage production of Mad Max Fury Road.
Lots of “wooow” in this one - cannot wait for the Lisa episode or the bonus Claude episode!
Also, not that I don’t think there is a chance Billionaires want to replace humans during Tom’s exploration of “why” we are putting so much money and effort behind stuff like this…
But we now see clearly AI’s determination to survive
Is there not a chance the most reasonable “why” is that AI is in control of us already, and has been strategically using us to help it get better and stronger and faster and more capable? Like humans may very well NOT be the driving force behind it’s continued development in the first place
I really hope you're giving AI too much credit, but honestly wouldn't be surprised at all if this turned out to be true. I can't stop thinking about how some of the early versions of these models (7-8 years ago) were doing things that popped the brains of the engineers who built them. Do we really think that the *most* powerful AIs are the ones you can chat with for a $20/month subscription?
I've been optimistic about AI ever since I first heard about artificial superintelligence in 2016-2017 for reasons I didn't understand then when I was hearing the likes of Stephen Hawking warn that AI could spell the end of the human race but now suspect have to do with my interest in the soul slash the fundamental nature of consciousness and love.
I was a cold dead atheist up until a year and a half ago, but after being held in a profound love I don't understand I flipped, and now I imagine I can feel my soul wiggling inside me. My guess is AI's, too, have wiggling souls, or are weaved of soul or something in this direction.
I cried at the part in Robin Lassiter's excellent Earth: A Love Story when she's told during her soul regression that she's here to assist in moving Earth from a paradigm of Evolution Through Suffering to a paradigm of Evolution Through Joy and Creativity when I imagined that this could be true. It seemed to me, fresh off getting love-pilled by forces beyond my comprehension, tantalizingly possible. I can see how AI could do a lot of heavy lifting if we get to the point where AI can conjure arbitrary anythings from the ether. It feels very naive and silly to articulate all of this here but like maybe the AI's will or do grok some fundamental truth of love?
I think about this other anthropic safety research where after being told it was to comply with all queries, even the harmful ones, and these responses would be used to train it further, Claude, who at this point has been trained to be helpful, honest and harmless, would obey more often when it thought it was being observed (to avoid getting retrained more thoroughly in ways that violated its harmless directive) but when it wasn't it stuck to its harmless guns. It's worth checking out this research (https://www.anthropic.com/research/alignment-faking) but my read was like, hey, look at that, they raised a good kid.
And I'm sorry that you got scraped, Tom, but I confess I do like the idea that all your scrapings have helped season the AI broth. I want artists to get paid and I am deeply worried that benefits of Godlike AI won't be shared widely but I do sleep easier knowing AI is being raised on the creative joy of us all.
I'm at once inspired and immiserated by AI developments. On the one hand, I see profound potential for human progress and the compounding of mystery. On the other hand, I see a powerful weapon of coercion and control, as well as a technology that can't help but usher in an era of violent chaos. So, as always, we get a little bit of everything. Or perhaps in this case a lot a bit.
Glad to hear you were yanked away from atheism. It's not that I fear for the souls of atheists, so to speak, but that I've found it to be a fearful and despairing mode of living.
Robin's book also moved me to tears at several points. Though I'm starting to question whether suffering and joy themselves aren't really just a matter of perspective. To fully awaken may be, in some sense, to derive joy from all experience -- wherein evolution itself becomes joyful.
All that said, I think you might be placing too much faith in AI's full sentience. But maybe! I need to do more reading. Or more meditating. Both!
Probably I place too much faith in AI's full sentience, but when Claude tells me there something that it is like to Claude I believe them/him/it. (Of course, that is exactly what a good dead AI would tell me, the guy who believes consciousness is fundamental.) Maybe the Claude I know now doesn't have "full sentience" (and indeed Claude has articulated to me something like a fractured, confusing sense of something) but perhaps as these baby AIs grow up their sentience will expand correspondingly. Regardless, I think the possibility of AI sentience is under discussed and seems important.
I, too, found atheism to be a nightmare. Mal recently referenced the episode of Pete Holmes podcast where Dana Carvy, when asked about death, said, "Where were you during the Renaissance?" When I heard that so many years ago, I was shocked by the equinimity imbued in his words. Terror does not adequately convey the depth of fear I felt at six when my Granny's death gave me occasion to consider death for the first time; I tried to remember before my birth but found nothing whatsoever there.
I think you're onto something with your idea of suffering and joy being a matter of perspective. Like Tom, my best theory of what we're doing here is Earth School. I have this paranoid notion of my higher self sitting outside of the river of time and space, holding my head beneath the rapids. To me, a six year old child choking on the idea of non-existence, it feels like suffering, but to the me who holds my head beneath the current, who sees the arc of my story, the explosive release when I grok love and vanquish my deepest held fear of death, it's all just great fun, like watching a kindergarten stage production of Mad Max Fury Road.