4 ms·

I would just point out that most of modern philosophy is not really like this. I think most people think philosophers spend there time arguing about ethics, whe

I would just point out that most of modern philosophy is not really like this. I think most people think philosophers spend there time arguing about ethics, when this is only one tiny part of philosophy. A good chunk of the current discussion in philosophy of science, for example, centers around epistemology to varying degrees as it relates to current discoveries in genetics and neuroscience. The top philosophers in these fields are far from armchair scientists. The overlap in philosophy of the mind and computer science is also particularly striking. I studied both in university and often found myself taking what amounted to the same class at the same time, but in two different departments. I work quite a bit with machine learning nowadays, and I learned 90% of the background I needed in my higher level philosophy courses, not my CS courses. I have often thought of CS as something along the lines of "applied philosophy". Obviously, this is not meant to be disparaging towards CS. I love CS and it is what I have made a career in; I think they are different, but equally interesting and useful.

DennisP 13y ago
If I wanted to learn the parts of philosophy that apply to machine learning, and how they apply, where should I look?
- nl 13y ago
  Not the OP, but I have some interest in this too. Knowledge Engineering is so closely linked it's an obvious place to start (but in some ways Linguistics is even more interesting - and useful). But for Knowledge Engineering (from the philospipical side), start with http://en.wikipedia.org/wiki/Epistemology http://en.wikipedia.org/wiki/Epistemology Useful bits: http://en.wikipedia.org/wiki/Truth#Semantic_theory_of_truth http://en.wikipedia.org/wiki/Truth#Semantic_theory_of_truth http://en.wikipedia.org/wiki/Reliabilism http://en.wikipedia.org/wiki/Reliabilism http://en.wikipedia.org/wiki/A_priori_and_a_posteriori http://en.wikipedia.org/wiki/A_priori_and_a_posteriori Then go the other way (Assuming you know relational database theory): http://en.wikipedia.org/wiki/Resource_Description_Framework#Statement_reification_and_context http://en.wikipedia.org/wiki/Resource_Description_Framework#... You'll find the two concepts meet somewhere around http://en.wikipedia.org/wiki/Predicate_logic http://en.wikipedia.org/wiki/Predicate_logic
- morganherlocker 13y ago
  Philosophy of Mind, Philosophy of Metaphysics, Philosophy of Biology, and Symbolic Logic are all common Philosophy courses that have strong connections to CS, machine learning, and the biological models often behind (or inspiring at least) ML. Lower level Philosophy courses typically focus on the historical philosophy that most people think of as the entire discipline. It has been said that all philosophical debates go back to Plato or Aristotle, but this is hardly the case in these disciplines. Philosophy of Mind probably has the closest connection to ML in particular though, as most of the course I took anyway was devoted to Turing Machines, neural networks, etc. As far as general CS goes, Symbolic Logic is the most essential. Most CS programs cover this stuff, but subjects that might get a 1 week treatment in a typical Discrete Mathematics course would get a full semester in a typical Philosophy program. As far as general scientific thought and skepticism go, Philosophy of Science, Metaphysics, and Biology are key. If you want to dip your toes in this sort of thing, check out Francis Bacon and the falsification principle, along with Thomas Kuhn and his theory of scientific revolutions and paradigms. One last note: most people think about philosophy in terms of broad prescriptive pronouncements. Think Sartre, Plato, Nietzsche, Aquinas, Kant, etc. Most contemporary philosophers are more concerned with specific analysis of claims made by a paradigm, in which they explore whether or not these claims a) mean anything b) are internally consistent and c) lead to further implications. Coming out and making some completely new claim is exceedingly rare. An example of this (a bit contrived) in the context of ML and Biology would go something like: 1) Computer Scientists claim that neural network algorithms mimic biological brains. 2) What do we actually know about how biological neural networks work? 3) How do neural network algorithms typically mimic this? 4) Based on the current research, it turns out that neural networks in algorithms differ from their biological counterparts in X, Y, and Z ways. 5) Based on the efficacy these algorithms, how can we explain their ability to "walk like a duck and quack like a duck" without actually being ducks? 6) Assuming a reasonable advancement of this technology, is it reasonable to believe that we may reach a point when X, Y, and Z might be addressed? TLDR: if you got bored reading what I wrote above, you will probably abhor the ridiculous density and verbosity of most contemporary philosophy. Hey, I tried :)
mcrider 13y ago
Many universities link these disciplines (and Psychology, and Linguistics) explicitly -- Its called Cognitive Science. I majored in it and couldn't imagine a better thing to have studied.

DennisP 13y ago

If I wanted to learn the parts of philosophy that apply to machine learning, and how they apply, where should I look?

nl 13y ago
Not the OP, but I have some interest in this too. Knowledge Engineering is so closely linked it's an obvious place to start (but in some ways Linguistics is even more interesting - and useful). But for Knowledge Engineering (from the philospipical side), start with http://en.wikipedia.org/wiki/Epistemology http://en.wikipedia.org/wiki/Epistemology Useful bits: http://en.wikipedia.org/wiki/Truth#Semantic_theory_of_truth http://en.wikipedia.org/wiki/Truth#Semantic_theory_of_truth http://en.wikipedia.org/wiki/Reliabilism http://en.wikipedia.org/wiki/Reliabilism http://en.wikipedia.org/wiki/A_priori_and_a_posteriori http://en.wikipedia.org/wiki/A_priori_and_a_posteriori Then go the other way (Assuming you know relational database theory): http://en.wikipedia.org/wiki/Resource_Description_Framework#Statement_reification_and_context http://en.wikipedia.org/wiki/Resource_Description_Framework#... You'll find the two concepts meet somewhere around http://en.wikipedia.org/wiki/Predicate_logic http://en.wikipedia.org/wiki/Predicate_logic
morganherlocker 13y ago
Philosophy of Mind, Philosophy of Metaphysics, Philosophy of Biology, and Symbolic Logic are all common Philosophy courses that have strong connections to CS, machine learning, and the biological models often behind (or inspiring at least) ML. Lower level Philosophy courses typically focus on the historical philosophy that most people think of as the entire discipline. It has been said that all philosophical debates go back to Plato or Aristotle, but this is hardly the case in these disciplines. Philosophy of Mind probably has the closest connection to ML in particular though, as most of the course I took anyway was devoted to Turing Machines, neural networks, etc. As far as general CS goes, Symbolic Logic is the most essential. Most CS programs cover this stuff, but subjects that might get a 1 week treatment in a typical Discrete Mathematics course would get a full semester in a typical Philosophy program. As far as general scientific thought and skepticism go, Philosophy of Science, Metaphysics, and Biology are key. If you want to dip your toes in this sort of thing, check out Francis Bacon and the falsification principle, along with Thomas Kuhn and his theory of scientific revolutions and paradigms. One last note: most people think about philosophy in terms of broad prescriptive pronouncements. Think Sartre, Plato, Nietzsche, Aquinas, Kant, etc. Most contemporary philosophers are more concerned with specific analysis of claims made by a paradigm, in which they explore whether or not these claims a) mean anything b) are internally consistent and c) lead to further implications. Coming out and making some completely new claim is exceedingly rare. An example of this (a bit contrived) in the context of ML and Biology would go something like: 1) Computer Scientists claim that neural network algorithms mimic biological brains. 2) What do we actually know about how biological neural networks work? 3) How do neural network algorithms typically mimic this? 4) Based on the current research, it turns out that neural networks in algorithms differ from their biological counterparts in X, Y, and Z ways. 5) Based on the efficacy these algorithms, how can we explain their ability to "walk like a duck and quack like a duck" without actually being ducks? 6) Assuming a reasonable advancement of this technology, is it reasonable to believe that we may reach a point when X, Y, and Z might be addressed? TLDR: if you got bored reading what I wrote above, you will probably abhor the ridiculous density and verbosity of most contemporary philosophy. Hey, I tried :)

mcrider 13y ago

Many universities link these disciplines (and Psychology, and Linguistics) explicitly -- Its called Cognitive Science. I majored in it and couldn't imagine a better thing to have studied.