5 ms·

Exactly. The OP has a very parochial view of philosophy, and what philosophers study.

visarga 8y ago
Please point me to a paper treating consciousness from the point of view of agent, environment and rewards (reinforcement learning). I have searched and found very few papers taking this angle. I concluded that philosophers don't know about it. In my view even using the word 'consciousness' is a bad thing because it is not well defined and has too much baggage (some people relate consciousness to spirit or soul, for example). Instead we should use different concepts, the ones I enumerate above. They cover sensing, prediction of future rewards (emotion), acting and learning. All are concrete and well defined, with possible implementations in AI.
- Emma_Goldman 8y ago
  I'm not sure what you're getting at here. What you say sounds like a perfectly interesting starting point for thinking about consciousness (although I'm not sure how it's supposed to fit with the fact that we're not computers, and have a distinct cognitive makeup, that mainstream psychology - 'cognitive psychology' - has committed itself to studying for several decades, with great profit). You would, in any case, have to specify what you mean by 'agent, environment, reward' - that sounds like a great deal of psychology. But what I took objection to was the suggestion that philosophy is or should be primarily interested in clearing up some of the conceptual problems attendant to contemporary computer science and engineering. I'm sure that's one interesting area of philosophy, but it is only one, modest area. There is the philosophy of aesthetics and literature, of ethics and politics, of religion and science, of epistemology and knowledge, of the philosophy of history and the history of philosophy, and so on. From what you said about existing philosophy ('Chinese room', qualia and 'hard problem' of consciousness) and the philosophy you would like to see (as doing the same thing as AI but worse), you completely discarded what the vast majority of what philosophers - now, and most certainly in the past - have actually done and been interested in. That is why I said you had a parochial view of philosophy. All that aside: I would be interested to hear how you think AI should change the philosophy of language and meaning.
  - visarga 8y ago
    The notions of agent, environment and reward are encountered in Reinforcement Learning which is a sub-field of ML but also relevant to biological agents. There is a great RL course on YouTube by David Silver (one of the creators of AlphaGo, which is probably the most famous application of RL so far). https://www.youtube.com/watch?v=2pWv7GOvuf0 https://www.youtube.com/watch?v=2pWv7GOvuf0 The course can properly set the perspective of RL and shed new light on the philosophy of mind, if you take what it says and then extrapolate to humans and other agents. What I find fascinating about RL is that it can be defined concretely. Consciousness can only be defined by reference to other words, in a less exact and concrete way. RL can also explain how meaning appears, based on future reward prediction. The rich sensations we have can be explained by an encoding-decoding architecture based on reconstruction error. Many difficulties in RL map back to difficulties humans have in choosing how to act - the exploration-exploitation tradeoff, instinct vs reason (two different ways to perform RL, one based purely on rewards and the other based on an internal model of the world and rewards). Some of the problems related to multi-agent RL are also covered in Game Theory, such as the prisoner's dilemma. Regarding philosophy of language: we have today numerical representations of meaning in words and phrases. They are usually represented as vectors in high dimensional space or sets of vectors. For example it is customary to use 300-1000 dimensions for representing words. These vectors have a nice property - the closer two words are in meaning, the smaller the angle between the two vectors. They are derived by trying to predict a word by the context, or vice versa, on a corpus of text several Gb in size. Many ideas from the philosophy of language, such as the meaning of words being related to the 'game' being played (activity with a purpose), emerge naturally from successful AI models. I'd say that where philosophy had a glimpse, AI has a testable implementation that can solve real world problems. Where philosophy uses mere words, AI uses probability distributions and datasets to define such models. The brain is probably doing something similar. Other things that AI has managed to do so far: to encode images into latent representations and back, to synthesise images. Same for speech - we have speech recognition and text to speech. Some modules used in neural networks are analogous to imagination, attention, memory, emotion, intuition and many other aspects of the mind. On narrow domains computers can already best humans at perception. The piece that is missing in AI to match human level is the prior knowledge encoded in our genes. We have been optimised to learn and function well in our environment by evolution. That means our verbal areas in the brain have a notion of invariance to time translation, visual areas have invariance to space translation, and conceptual areas have an invariance to permutation. There might be more invariances but we just don't know yet and that's why AI models are not yet up to par with humans. But we can still learn a lot about ourselves by analogy with AI agents. And that's where I think philosophy should listen.
- foldr 8y ago
  https://scholar.google.co.uk/scholar?hl=en&as_sdt=0%2C5&q=behaviorism+conciouscness&btnG= https://scholar.google.co.uk/scholar?hl=en&as_sdt=0%2C5&q=be...
- wwweston 8y ago
  The question about which concepts you should use and why (and the arguments supporting them) might be a decent philosophical one. "Too much baggage" isn't particularly well defined, though, nor is it a strong argument to say that some people relate consciousness to spirit or soul (some people also use terms like "momentum" or "norm" in imprecise ways but it doesn't keep practitioners for whom those terms have precise well-defined meanings from deploying them effectively). "we should use different concepts... All are concrete and well defined" Or perhaps we should use concepts from clockwork engineering or mill machinery. Very concrete and well defined.

Please point me to a paper treating consciousness from the point of view of agent, environment and rewards (reinforcement learning). I have searched and found very few papers taking this angle. I concluded that philosophers don't know about it. In my view even using the word 'consciousness' is a bad thing because it is not well defined and has too much baggage (some people relate consciousness to spirit or soul, for example). Instead we should use different concepts, the ones I enumerate above. They cover sensing, prediction of future rewards (emotion), acting and learning. All are concrete and well defined, with possible implementations in AI.

Emma_Goldman 8y ago
I'm not sure what you're getting at here. What you say sounds like a perfectly interesting starting point for thinking about consciousness (although I'm not sure how it's supposed to fit with the fact that we're not computers, and have a distinct cognitive makeup, that mainstream psychology - 'cognitive psychology' - has committed itself to studying for several decades, with great profit). You would, in any case, have to specify what you mean by 'agent, environment, reward' - that sounds like a great deal of psychology. But what I took objection to was the suggestion that philosophy is or should be primarily interested in clearing up some of the conceptual problems attendant to contemporary computer science and engineering. I'm sure that's one interesting area of philosophy, but it is only one, modest area. There is the philosophy of aesthetics and literature, of ethics and politics, of religion and science, of epistemology and knowledge, of the philosophy of history and the history of philosophy, and so on. From what you said about existing philosophy ('Chinese room', qualia and 'hard problem' of consciousness) and the philosophy you would like to see (as doing the same thing as AI but worse), you completely discarded what the vast majority of what philosophers - now, and most certainly in the past - have actually done and been interested in. That is why I said you had a parochial view of philosophy. All that aside: I would be interested to hear how you think AI should change the philosophy of language and meaning.
- visarga 8y ago
  The notions of agent, environment and reward are encountered in Reinforcement Learning which is a sub-field of ML but also relevant to biological agents. There is a great RL course on YouTube by David Silver (one of the creators of AlphaGo, which is probably the most famous application of RL so far). https://www.youtube.com/watch?v=2pWv7GOvuf0 https://www.youtube.com/watch?v=2pWv7GOvuf0 The course can properly set the perspective of RL and shed new light on the philosophy of mind, if you take what it says and then extrapolate to humans and other agents. What I find fascinating about RL is that it can be defined concretely. Consciousness can only be defined by reference to other words, in a less exact and concrete way. RL can also explain how meaning appears, based on future reward prediction. The rich sensations we have can be explained by an encoding-decoding architecture based on reconstruction error. Many difficulties in RL map back to difficulties humans have in choosing how to act - the exploration-exploitation tradeoff, instinct vs reason (two different ways to perform RL, one based purely on rewards and the other based on an internal model of the world and rewards). Some of the problems related to multi-agent RL are also covered in Game Theory, such as the prisoner's dilemma. Regarding philosophy of language: we have today numerical representations of meaning in words and phrases. They are usually represented as vectors in high dimensional space or sets of vectors. For example it is customary to use 300-1000 dimensions for representing words. These vectors have a nice property - the closer two words are in meaning, the smaller the angle between the two vectors. They are derived by trying to predict a word by the context, or vice versa, on a corpus of text several Gb in size. Many ideas from the philosophy of language, such as the meaning of words being related to the 'game' being played (activity with a purpose), emerge naturally from successful AI models. I'd say that where philosophy had a glimpse, AI has a testable implementation that can solve real world problems. Where philosophy uses mere words, AI uses probability distributions and datasets to define such models. The brain is probably doing something similar. Other things that AI has managed to do so far: to encode images into latent representations and back, to synthesise images. Same for speech - we have speech recognition and text to speech. Some modules used in neural networks are analogous to imagination, attention, memory, emotion, intuition and many other aspects of the mind. On narrow domains computers can already best humans at perception. The piece that is missing in AI to match human level is the prior knowledge encoded in our genes. We have been optimised to learn and function well in our environment by evolution. That means our verbal areas in the brain have a notion of invariance to time translation, visual areas have invariance to space translation, and conceptual areas have an invariance to permutation. There might be more invariances but we just don't know yet and that's why AI models are not yet up to par with humans. But we can still learn a lot about ourselves by analogy with AI agents. And that's where I think philosophy should listen.
foldr 8y ago
https://scholar.google.co.uk/scholar?hl=en&as_sdt=0%2C5&q=behaviorism+conciouscness&btnG= https://scholar.google.co.uk/scholar?hl=en&as_sdt=0%2C5&q=be...
wwweston 8y ago
The question about which concepts you should use and why (and the arguments supporting them) might be a decent philosophical one. "Too much baggage" isn't particularly well defined, though, nor is it a strong argument to say that some people relate consciousness to spirit or soul (some people also use terms like "momentum" or "norm" in imprecise ways but it doesn't keep practitioners for whom those terms have precise well-defined meanings from deploying them effectively). "we should use different concepts... All are concrete and well defined" Or perhaps we should use concepts from clockwork engineering or mill machinery. Very concrete and well defined.