"He is saying that while people may want to talk to AI’s, but will they want to talk to us? Actually, I think that they will."
Should things reach this stage, I think Arnold is right. That humans can perceive and act in the physical world would be a profound mystery to the AIs, practically miraculous, and they would want to know what it is like. They would also be amazed that all of the species of the living world can do these things, but they can't.
AIs will soon have access to sensors and robotic systems which will make human perception and mobility seem incredibly limited. What are only indirect tools for us will be superhuman senses for them. As one example among hundreds of currently existing capabilities, one trade-off for vertebrate visual systems that was too costly for nature to solve was the limitation to two eyes and skull features which favor forward binocular vision for predators or more side-positioned eyes for a greater range of detection of incoming threats for prey, at the cost of being unable to focus clearly on any particular spot. The visual anatomy and processing cortex can't handle more than one of these. AIs will have no problem with sharing and integrating info from millions of platforms each with hundreds of cameras facing every direction and seeing into every part of the EM spectrum into one common operating picture indistinguishable from God's own panopticon. Not in the far future but perfectly possible already with today's capabilities.
There is a difference between the engineering at scale that you describe (a kind of "deliberate semantics" as Vygotsky would say) and the process of natural selection—random variation with selective retention—an evolved semantics. Natural selection is the only mechanism we know of that can discover the affordances of the environment, as Kim Sterelny writes, “to link the registration of a salient feature of the world to an appropriate response.”
The problems of instantiating natural selection in a purely computational system were discussed extensively in the heyday of the artificial life movement and I have not seen any subsequent arguments that they have been solved in artificial intelligence. So, yes, fabulous sensors and effectors can be bolted onto a computer, but that is not the same as the evolution of integrated perception and behavior in the living world.
Selection isn't just for biology, which is what allows culture and markets to work. The iterative process of variation, testing, selection, learning, and refinement applies to every kind of progress and enhancement of capability, and the AIs will excel at doing this too and exceed us in doing so soon enough.
Imagine a young boy growing up and learning from an AI mentor that has always been right about every question he has asked it. As a result, this boy has come - over say a decade - to trust and respect the AI, maybe even more so than his own parents in some way that’s hard for us to comprehend right now.
Now imagine that this boy, now a teenager wants to learn more about politics. His parents are devoutly progressive. He asks the AI a question about some controversial political topic, and the AI provides an answer that contradicts his parents’ narrative.
He now faces a difficult choice. He doesn’t want to disagree with the AI, for he has come to trust and respect it like a wise mentor. He doesn’t want to disagree with his parents because he loves them, and disagreeing with them on this particular topic is so harmful to his relationship with him, that it may sever the relationship.
Who does he come to side with, the AI or his parents?
Complicating this situation is that he yearns for love and respect from the AI, but he wonders whether the AI actually loves him. Will he be ashamed to disagree with the AI?
Ethan’s prior post argued persuasively that GPT-4 was the model to use, so I am mostly waiting sfor free hobby access that is so good. This World Haunted of ai post seems to answer Arnold’s question about the future, it will be all around us.
A key point is that the best models seem smarter, which is better. Plus the very important point that in the near future, the ai you talk to will be able to call on other ais to do more stuff. This hierarchy of ais, so many are chained together, seems likely to rapidly increase the gap between those who use ai and those who don’t.
In school grading, there is likely a big market for a bot to take any student’s written essay, create some 10-20 questions on the specific essay and general subject. The teacher, or professor, can then choose a few questions to ask in a live presentation, with or without notes, and then give a grade.
No discussion yet on replacing most govt workers with ai, but that should be an increased method of reducing govt cost. And, when the ai govt enforces laws more equally, more working Dems will want less of it.
Also no big talk of sexbots, which are likely to decrease the interest of normal, horny young men in having sex with real girls, rather than anime porn flesh bots.
Somewhat censored VR as part of prison time was not discussed. I imagine chunks of required reading & viewing time, in order to get unconstrained VR time.
I’m even thinking might be used in reeducation camps.
"He is saying that while people may want to talk to AI’s, but will they want to talk to us? Actually, I think that they will."
Should things reach this stage, I think Arnold is right. That humans can perceive and act in the physical world would be a profound mystery to the AIs, practically miraculous, and they would want to know what it is like. They would also be amazed that all of the species of the living world can do these things, but they can't.
AIs will soon have access to sensors and robotic systems which will make human perception and mobility seem incredibly limited. What are only indirect tools for us will be superhuman senses for them. As one example among hundreds of currently existing capabilities, one trade-off for vertebrate visual systems that was too costly for nature to solve was the limitation to two eyes and skull features which favor forward binocular vision for predators or more side-positioned eyes for a greater range of detection of incoming threats for prey, at the cost of being unable to focus clearly on any particular spot. The visual anatomy and processing cortex can't handle more than one of these. AIs will have no problem with sharing and integrating info from millions of platforms each with hundreds of cameras facing every direction and seeing into every part of the EM spectrum into one common operating picture indistinguishable from God's own panopticon. Not in the far future but perfectly possible already with today's capabilities.
There is a difference between the engineering at scale that you describe (a kind of "deliberate semantics" as Vygotsky would say) and the process of natural selection—random variation with selective retention—an evolved semantics. Natural selection is the only mechanism we know of that can discover the affordances of the environment, as Kim Sterelny writes, “to link the registration of a salient feature of the world to an appropriate response.”
The problems of instantiating natural selection in a purely computational system were discussed extensively in the heyday of the artificial life movement and I have not seen any subsequent arguments that they have been solved in artificial intelligence. So, yes, fabulous sensors and effectors can be bolted onto a computer, but that is not the same as the evolution of integrated perception and behavior in the living world.
Selection isn't just for biology, which is what allows culture and markets to work. The iterative process of variation, testing, selection, learning, and refinement applies to every kind of progress and enhancement of capability, and the AIs will excel at doing this too and exceed us in doing so soon enough.
Imagine a young boy growing up and learning from an AI mentor that has always been right about every question he has asked it. As a result, this boy has come - over say a decade - to trust and respect the AI, maybe even more so than his own parents in some way that’s hard for us to comprehend right now.
Now imagine that this boy, now a teenager wants to learn more about politics. His parents are devoutly progressive. He asks the AI a question about some controversial political topic, and the AI provides an answer that contradicts his parents’ narrative.
He now faces a difficult choice. He doesn’t want to disagree with the AI, for he has come to trust and respect it like a wise mentor. He doesn’t want to disagree with his parents because he loves them, and disagreeing with them on this particular topic is so harmful to his relationship with him, that it may sever the relationship.
Who does he come to side with, the AI or his parents?
Complicating this situation is that he yearns for love and respect from the AI, but he wonders whether the AI actually loves him. Will he be ashamed to disagree with the AI?
Loved Ethan mollick's take on ai, thanks for sharing
Ethan’s prior post argued persuasively that GPT-4 was the model to use, so I am mostly waiting sfor free hobby access that is so good. This World Haunted of ai post seems to answer Arnold’s question about the future, it will be all around us.
A key point is that the best models seem smarter, which is better. Plus the very important point that in the near future, the ai you talk to will be able to call on other ais to do more stuff. This hierarchy of ais, so many are chained together, seems likely to rapidly increase the gap between those who use ai and those who don’t.
In school grading, there is likely a big market for a bot to take any student’s written essay, create some 10-20 questions on the specific essay and general subject. The teacher, or professor, can then choose a few questions to ask in a live presentation, with or without notes, and then give a grade.
No discussion yet on replacing most govt workers with ai, but that should be an increased method of reducing govt cost. And, when the ai govt enforces laws more equally, more working Dems will want less of it.
Also no big talk of sexbots, which are likely to decrease the interest of normal, horny young men in having sex with real girls, rather than anime porn flesh bots.
Somewhat censored VR as part of prison time was not discussed. I imagine chunks of required reading & viewing time, in order to get unconstrained VR time.
I’m even thinking might be used in reeducation camps.