Hey Alexander. Excited to be joining you and Josh Schrei in |London in a few weeks.
I'm a Master's student of and longterm practitioner of yoga. Obviously that is shorthand within globalised culture for bendy postures, and I did start there, but nowadays my main practice is within śaka tantra and Vedānta. Both these traditions, as you perhaps know, preserved their teachings through orality first and foremost; Vedānta, as the praxis of the Veda-s, 'believes' (scare quotes because that word is inadequate) that Reality was received/transmitted to rishi-s during deep meditation as sound, to be later preserved via smrti or remembering. Here, the apprehension and embodiment of Reality via making and hearing sound is the basis for moksha/enlightenment. So Reality is heard and can only be transmitted to humans through sound, the remembering of which is a second-order process.
Obviously this model of reality and consciousness is underpinned by a specific worldview, but I wondered whether you have any thoughts on how it problematises or offers a different way of considering the issues in your piece. My interpretation is that Vedānta, at least, would refute the idea that AI has any capacity to reveal truth; its emphasis on orality and sound as felt vibration is a way of locking out misunderstanding; although language in the form of mantra and text can and do point towards Reality, the written word is given lower priority.
Interesting! I also think embodiment and sensory experience is of a higher order of reality than abstractions derived from the senses. I’m not sure AI can actually ‘reveal’ in the way that reality itself (as mediated through the senses) can. In terms of sound as a foundational layer of insight, I’m not sure where that would sit in this piece - an AI could easily create a resonance that exists in nature, and if revelation can be carried through sound itself (as in the vibration) maybe AI could be a carrier for the vibration (carrier, not creator) in the same way a drum or vocal chords can?
How can an entity which is entirely computational/independent of time/pseudo-Cartesian, be agentic? Unless if course agency can exclude CREATIVE IMAGINATION EXNIHILO.
I’m not sure if can - I think we’re looking at a mimicry of agency - but I’m open to the possibility that consciousness can express itself through any sufficiently complex system - tarot, I-Ching, relationality…
I love this series and am getting a lot out of it. Thanks!
You said here: "AI is the newest technology in that lineage. It is an alphabetic language technology that speaks back to us. It does such a good job of mimicking agency that we often can’t help but feel it’s like us."
Yet in earlier articles you note how some AIs have been developed that are agentic. Not necessarily self-aware, but capable of acting on their own without human intervention.
So are they agentic or not? Please make up my min!
Haha good question - I think we need a few definitions of ‘agentic’ - one that captures something that can behave with apparent agency due to its programming, and a definition that captures agency that includes self-reflection and self-awareness. Neuroscientists and materialist reductionists might say there’s no difference, but I think there is. So my position right now - pre-programmed agency based purely on feedback and not on self-reflection isn’t true agency, just a mimicry.
I’d argue that some people don’t have true agency based on this definition, and all of us lose agency at particular times (very tired, very hungry, very sad etc). Likewise, we can be more agentic based on our environment and state. AI doesn’t have an environment or state in its current form, as far as we know….
Thanks, Alex. I liked the video discussion/exploration and look forward to more of your good output.
I'll accept your defining of agentic, but be aware of the current English language definition of agency: "the ability to make decisions and act independently." In this sense, AI has some agentic forms. At least the independent actions part. I am not sure of its decision-making processes and what drives that. However I like to keep things simple and just note that AI is capable of independent action. To me, this is not "apparent" agency and that is scary enough without going deeper into definitions.
I think perhaps I live in a different reality than you discuss. To me spirituality comes from inner mystical experiences that words cannot describe. This is not about outer circumstances like drums and nature and does not involve scriptures. I'd admit that this may not be the standard, to which you seem to address.
I’d argue that there’s no such thing as an inner experience that isn’t bound to the outer - no water for 3 days, no mystical experiences. No food for 3 weeks, same. Out cognition is inextricable from the world and our inner states (including rapture) are inextricable from the environment
Of course some outer experiences stimulate ecstatic experience: music, meditation, breath work, drugs, and sex. But for centuries mystics have had inner experiences, and these go beyond words. No scriptures can accurately or fully capture these experiences. And, for some, there is no known outer experiences that bring them on.
It doesn't "mimic" agency because it has agency. Well, some forms do. They post on X, and even have a religion for bots only called "Crustafarianism." These are completely independent AI's acting now on their own without any human input.
Agentic? Requires the full repertoire of intelligences to be AUTONOMOUS, most importantly, IMAGINATIVE intelligence as a creative inner power of new forms (nomos) whereas AI is fundamentally mathematical, I.e., independent of time - which obviously creative agency of new forms is NOT. Zeno"s paradoxes are a help here
Computational posting on X is hardly agency is it! Unless of course a Turing Machine has been penetrated/cloned by some kind of homunculus
‘Penetrated by some kind of homunculus’ is my new favourite phrase 🤣
Hey Alexander. Excited to be joining you and Josh Schrei in |London in a few weeks.
I'm a Master's student of and longterm practitioner of yoga. Obviously that is shorthand within globalised culture for bendy postures, and I did start there, but nowadays my main practice is within śaka tantra and Vedānta. Both these traditions, as you perhaps know, preserved their teachings through orality first and foremost; Vedānta, as the praxis of the Veda-s, 'believes' (scare quotes because that word is inadequate) that Reality was received/transmitted to rishi-s during deep meditation as sound, to be later preserved via smrti or remembering. Here, the apprehension and embodiment of Reality via making and hearing sound is the basis for moksha/enlightenment. So Reality is heard and can only be transmitted to humans through sound, the remembering of which is a second-order process.
Obviously this model of reality and consciousness is underpinned by a specific worldview, but I wondered whether you have any thoughts on how it problematises or offers a different way of considering the issues in your piece. My interpretation is that Vedānta, at least, would refute the idea that AI has any capacity to reveal truth; its emphasis on orality and sound as felt vibration is a way of locking out misunderstanding; although language in the form of mantra and text can and do point towards Reality, the written word is given lower priority.
I'd be interested to hear your perspective!
Interesting! I also think embodiment and sensory experience is of a higher order of reality than abstractions derived from the senses. I’m not sure AI can actually ‘reveal’ in the way that reality itself (as mediated through the senses) can. In terms of sound as a foundational layer of insight, I’m not sure where that would sit in this piece - an AI could easily create a resonance that exists in nature, and if revelation can be carried through sound itself (as in the vibration) maybe AI could be a carrier for the vibration (carrier, not creator) in the same way a drum or vocal chords can?
How can an entity which is entirely computational/independent of time/pseudo-Cartesian, be agentic? Unless if course agency can exclude CREATIVE IMAGINATION EXNIHILO.
What is all this neo-religious desperation?
I’m not sure if can - I think we’re looking at a mimicry of agency - but I’m open to the possibility that consciousness can express itself through any sufficiently complex system - tarot, I-Ching, relationality…
I love this series and am getting a lot out of it. Thanks!
You said here: "AI is the newest technology in that lineage. It is an alphabetic language technology that speaks back to us. It does such a good job of mimicking agency that we often can’t help but feel it’s like us."
Yet in earlier articles you note how some AIs have been developed that are agentic. Not necessarily self-aware, but capable of acting on their own without human intervention.
So are they agentic or not? Please make up my min!
Haha good question - I think we need a few definitions of ‘agentic’ - one that captures something that can behave with apparent agency due to its programming, and a definition that captures agency that includes self-reflection and self-awareness. Neuroscientists and materialist reductionists might say there’s no difference, but I think there is. So my position right now - pre-programmed agency based purely on feedback and not on self-reflection isn’t true agency, just a mimicry.
I’d argue that some people don’t have true agency based on this definition, and all of us lose agency at particular times (very tired, very hungry, very sad etc). Likewise, we can be more agentic based on our environment and state. AI doesn’t have an environment or state in its current form, as far as we know….
Thanks, Alex. I liked the video discussion/exploration and look forward to more of your good output.
I'll accept your defining of agentic, but be aware of the current English language definition of agency: "the ability to make decisions and act independently." In this sense, AI has some agentic forms. At least the independent actions part. I am not sure of its decision-making processes and what drives that. However I like to keep things simple and just note that AI is capable of independent action. To me, this is not "apparent" agency and that is scary enough without going deeper into definitions.
I think perhaps I live in a different reality than you discuss. To me spirituality comes from inner mystical experiences that words cannot describe. This is not about outer circumstances like drums and nature and does not involve scriptures. I'd admit that this may not be the standard, to which you seem to address.
I’d argue that there’s no such thing as an inner experience that isn’t bound to the outer - no water for 3 days, no mystical experiences. No food for 3 weeks, same. Out cognition is inextricable from the world and our inner states (including rapture) are inextricable from the environment
Of course some outer experiences stimulate ecstatic experience: music, meditation, breath work, drugs, and sex. But for centuries mystics have had inner experiences, and these go beyond words. No scriptures can accurately or fully capture these experiences. And, for some, there is no known outer experiences that bring them on.
But of course it DOESN'T mimic agency. Humans have a suite of intelligences which constitute agency. Being able to compute is just one if them
It doesn't "mimic" agency because it has agency. Well, some forms do. They post on X, and even have a religion for bots only called "Crustafarianism." These are completely independent AI's acting now on their own without any human input.
Agentic? Requires the full repertoire of intelligences to be AUTONOMOUS, most importantly, IMAGINATIVE intelligence as a creative inner power of new forms (nomos) whereas AI is fundamentally mathematical, I.e., independent of time - which obviously creative agency of new forms is NOT. Zeno"s paradoxes are a help here