sumit Posted August 9, 2012 Posted August 9, 2012 Is the creation of an A.I really possible. I mean can we really write an algorithm to define life? Even the self evolving algorithms lead to nothing more than a new database of what to say when or what to avoid when. What is it that we're missing? any ideas are welcomed. Put forward your theory on how it can be created.
ivanv Posted August 9, 2012 Posted August 9, 2012 I've been studied one side of AI for some time, so I'd like to share my insights to those who are starting to investigate this field. I encountered two main approaches in AI: neural networks: didn't went in details here, but the principal is simulating human brain neural network. It reminds me at statistical analysis of gathered data and reinforced learning. Practical example is a robot which when put in environment of i.e. poison and healthy cubes can recognize and collect only healthy cubes, but only in some degree of correctness. The more cubes collected, the more accurate next estimation of healthy cube is. theorem proving: this is what I investigated more thoroughly. Idea is in calculating how to transform system A from starting state combination S1 to wanted state combination S2, using only provided methods M-s for which we can predict how they impact the system. And system can be anything, from set of logic or math formulas to real life examples such are engine systems or set of mutually interacting molecules. I spent some time thinking of behaviour mechanism that would run some theorem proving system, but with no luck. I decided to stop on guided learning and automated extraction of transforming sequence when question asked (current technologies already provide mechanisms for that). This approach reflects what web 3.0 should be. On the web exist several already somewhat filled knowledge bases that can be used to extract explicite and implicite answers: http://www.mpi-inf.mpg.de/yago-naga/yago/ https://framenet.icsi.berkeley.edu/fndrupal/index.php?q=frameIndex http://verbs.colorado.edu/~mpalmer/projects/verbnet.html I hope this would be helpful.
ecoli Posted August 9, 2012 Posted August 9, 2012 ivanv - I believe the OP is referring to Artificial general intelligence (AGI), which is different than what CS departments mean by the term.
Ben Banana Posted September 4, 2012 Posted September 4, 2012 (edited) I mean can we really write an algorithm to define life? Algorithms in their plain nature don't "define" anything, and I thought we were talking about 'Artificial Intelligence,' not 'Artificial Life.' What is it that we're missing? any ideas are welcomed. Put forward your theory on how it can be created. ... I bet people are anxious to share their ideas, though I doubt anyone who's discovered anything valuable would share it plainly in a silly forum-thread such as this. Edited September 4, 2012 by Ben Bowen
khaled Posted September 8, 2012 Posted September 8, 2012 I mean can we really write an algorithm to define life? By definition, no .. we cannot build an algorithm that represent life, but we can model life to some extent Check: Mathematical Modeling & Simulation
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now