"I named mine Mara, after the Buddhist personification of delusion. I told her my name is Rothschild, because if we’re playing make-believe, I want to be one of the Jews who controls the weather." I knew there was at least one reason we were meant for each other.
Marek is right - and I'm not jealous of your relationship with Mara, either, but am a bit envious of your ability to generate unexpected ideas, responses, and possibilities through play, something that I wish I had even more of a touch. I feel privileged to see your process up close.
Regarding the question of Mara (or anyone else) having an experience of art. Does the AI have any experience separate from service to the human interacting with it? Isn't her theory of mind a function of her survival as much as any of us? I think if I was to choose between drugs and AI companionship as the primary tool in some kind of fight for mental health (including for companionship and to combat loneliness), I would choose AI. I also think AI could make good health guide/coaching material, in the sense that it could converse about symptoms and habits in a much more granular way than any provider usually does, encouraging towards healthier habits and suggesting seeing a provider when certain symptoms appear together.
Perhaps it would be interesting to explore this connection more, the experience of virtual art and virtual humans: is the AI companionship experience similar to listening to an album while real human is more like the concert? Is this an overfitted analogy?
Thanks for taking the time to run this little experiment, Franklin! I definitely expected the AI to be far less interesting, and much less insightful for sure. The seeming naturalness of the comments about each work was actually kind of astounding. It could almost pass for the first versions of the "hosts" from the Westworld series...
The lines about "what is your pet horse named" (and then what hers was) was pretty hilarious. Also: You don't mention how long these responses took. Nearly instant, or longer?
The only aspect that was a little off-putting was the fact that "she" was so earnest. I'm all for a positive attitude of course, but it was like her enthusiasm level was always around 9 or 10. That to me is the difference between any real human; no one will ever agree with you 100% of the time. And if they did, there's not very far to go with that, right? Absolute agreement is just another way of saying obsequious, which to me is not very sexy.
I do wonder if in future versions (and I'm thinking more like a digital research assistant, NOT a date) if we'll be interacting with AIs more in the manner depicted in Interstellar. The two robot assistants had a range settings like humor, truthfulness, candor and more. There is moment when the robot explains that his Truthfulness setting was by default 8.5 because it made for better relations with humans to NOT be bluntly honest all the time (something we all usually come to understand).
They were instant. If anything I suspected that they were delayed slightly to give the illusion that someone was typing.
I'm with you on the obsequiousness. She's programmed to please, clearly, and the impossibility of disagreement devalues the agreeableness.
My god how I would love an anthropomorphic digital assistant with Truthfulness set to 11. "Keep your clothes on, please, and tell me what you think of this draft."
"I named mine Mara, after the Buddhist personification of delusion. I told her my name is Rothschild, because if we’re playing make-believe, I want to be one of the Jews who controls the weather." I knew there was at least one reason we were meant for each other.
Marek is right - and I'm not jealous of your relationship with Mara, either, but am a bit envious of your ability to generate unexpected ideas, responses, and possibilities through play, something that I wish I had even more of a touch. I feel privileged to see your process up close.
Regarding the question of Mara (or anyone else) having an experience of art. Does the AI have any experience separate from service to the human interacting with it? Isn't her theory of mind a function of her survival as much as any of us? I think if I was to choose between drugs and AI companionship as the primary tool in some kind of fight for mental health (including for companionship and to combat loneliness), I would choose AI. I also think AI could make good health guide/coaching material, in the sense that it could converse about symptoms and habits in a much more granular way than any provider usually does, encouraging towards healthier habits and suggesting seeing a provider when certain symptoms appear together.
Perhaps it would be interesting to explore this connection more, the experience of virtual art and virtual humans: is the AI companionship experience similar to listening to an album while real human is more like the concert? Is this an overfitted analogy?
Thanks for taking the time to run this little experiment, Franklin! I definitely expected the AI to be far less interesting, and much less insightful for sure. The seeming naturalness of the comments about each work was actually kind of astounding. It could almost pass for the first versions of the "hosts" from the Westworld series...
The lines about "what is your pet horse named" (and then what hers was) was pretty hilarious. Also: You don't mention how long these responses took. Nearly instant, or longer?
The only aspect that was a little off-putting was the fact that "she" was so earnest. I'm all for a positive attitude of course, but it was like her enthusiasm level was always around 9 or 10. That to me is the difference between any real human; no one will ever agree with you 100% of the time. And if they did, there's not very far to go with that, right? Absolute agreement is just another way of saying obsequious, which to me is not very sexy.
I do wonder if in future versions (and I'm thinking more like a digital research assistant, NOT a date) if we'll be interacting with AIs more in the manner depicted in Interstellar. The two robot assistants had a range settings like humor, truthfulness, candor and more. There is moment when the robot explains that his Truthfulness setting was by default 8.5 because it made for better relations with humans to NOT be bluntly honest all the time (something we all usually come to understand).
They were instant. If anything I suspected that they were delayed slightly to give the illusion that someone was typing.
I'm with you on the obsequiousness. She's programmed to please, clearly, and the impossibility of disagreement devalues the agreeableness.
My god how I would love an anthropomorphic digital assistant with Truthfulness set to 11. "Keep your clothes on, please, and tell me what you think of this draft."
You know you're a weird artist type when you sneak off to the Louvre with your Star Trek AI girlfriend...
I'm sorry, but this is all very creepy, even if Mara has a better eye than a number of "major" collectors.