Apologies to those of you who will see this announcement twice. The time has come! Calling all Haskell programmers interested in AI! I've established a new home base at http://www.haskell.org/haskellwiki/AI . Functional programming has long been recognized as an excellent paradigm for Artificial Intelligence. It's time to make Haskell an important language in that area! The goal is to provide a toolkit of AI tools in Haskell that can easily be applied to a wide variety of applications: e.g., Neural Networks, genetic algorithms, natural language processing tools, etc.
Andrew Wagner wrote:
Functional programming has long been recognized as an excellent paradigm for Artificial Intelligence. One reason why LISP is used for AI is (in my opinion, more important than functional programming) that ist's easy to work with "symbols" in LISP. How easy is this in Haskell? Does it matter that, in Haskell, every symbol must have a type that's known at compile time?
Sebastian Setzer
I think that more important than dynamic typing is the high level abstraction provided by functional programming. And there, Haskell is certainly lisp's equal, if not its superior. In fact, I would argue that it's superior, due to lazy evaluation, and all its implications (monads, infinite streams, etc.). On 3/28/07, Setzer, Sebastian (ext) <sebastian.setzer.ext@siemens.com> wrote:
Andrew Wagner wrote:
Functional programming has long been recognized as an excellent paradigm for Artificial Intelligence. One reason why LISP is used for AI is (in my opinion, more important than functional programming) that ist's easy to work with "symbols" in LISP. How easy is this in Haskell? Does it matter that, in Haskell, every symbol must have a type that's known at compile time?
Sebastian Setzer _______________________________________________ Haskell mailing list Haskell@haskell.org http://www.haskell.org/mailman/listinfo/haskell
participants (2)
-
Andrew Wagner -
Setzer, Sebastian (ext)