Artificial intelligence (AI) is the intelligence exhibited by machines or software. It is an academic field of study which studies the goal of creating intelligence.
Background
However the current AI scenario is not like what I said. Modeling a computer system which is able to act like us, think like us and crack complex problems is not possible unless we fully understand how does our BRAIN functions. In spite of the hurdles we have had many successful stories viz., Deep Blue defeated Garry Kasparov in 1997, Watson computer defeated a human in a quiz competition, etc.
There is a mix of ideas around which many applications are now being build that technically aren't AI such as face recognition ( yes I call it automation NOT AI ) , big data analysis, etc. Because these ideas deal with logical actions based up on the program it was built for and not with the "deep learning" or "machine learning" or "conscious thinking" technique which is what AI means. Hence there is a prominent line of difference between automation and AI which we need to understand.
The area of "deep learning" that makes the computer more cognitive and adaptive is my area of interest. I think automation (eg Internet of Things) which is now becoming the next big thing now would finally make space for true AI the next big thing in future.
Application
A smart application of true AI would be to model our conscious mind and store it in the soft form! :D Yes . Imagine you are not alive anymore but your conscious mind is still there to interact with others. It can analyse, think of a problem and crack it just like you did. Every bit of information, ideas, visions of your life that you thought would be beneficial for humans will be available to the humans through it. In short your brain will live forever! Sounds just like the science fiction, isn't it? Remember Man of Steal where Kal-El could interact with his father Jor-El after his death?
Project Idea
A computer model which can mimic us and our thinking is very much possible to create with our present understanding of AI. We have seen successful creation of intelligent software like Siri and Watson which are getting more intelligent day by day. To start off one can think of modelling a simple system that can analyse small change is data sets and then adept to that change automatically. Or simply we can build a simple short code that can arrive to a solution without programming it to do the same.
For eg.
INPUT - Arriving to a value of 16 by any method from any two integers ;
OUTPUT - f1(X,Y) , f2(X,Y) , f2(X,Y),...
f1, f2, f3, .. are small codes that solves a problem to arrive a value of 16 from different sets of X and Ys. But each methods f1, f2, f3,.. of arriving to a value 16 are different.
The next step would be to write a software that can code an efficient algorithm from a given problem set. With these we can gradually learn how to train a computer system that reacts to different kind of situations. Finally we will be able to think of a system which can make intelligent fundamental decisions, ushering a path to our goal.
I found a blog article listing some great introductory AI courses from available coursewares like Edx, MIT courseware, Udacity, etc. here. These courses are great for beginners. If you think there are other great AI open courses available , add them in the comments. I will add them here.
While there are stripes relevant AI disciplines that normally people use to start with viz. control theory, linguistics, computer engineering, neuroscience, etc. One should directly start off with pragmatic AI project ideas without going through these academic approaches. I have my personal experience of learning NLP ( Natural Language Processing ) which concerns with the linguistic intelligence. I found that the method of learning all the tools of NLP and then going for a project is too boring. Instead one should start off with their own ideas and refer relevant books or tools to make prototypes.
Therefore my humble request to all the hackers, makers and developers is to take the next step towards AI and to all idea developers to make their ideas open source. Because I think your ideas aren't alive or worth it unless you make them happen. I just started to make all my ideas open source. :)
A computer model which can mimic us and our thinking is very much possible to create with our present understanding of AI. We have seen successful creation of intelligent software like Siri and Watson which are getting more intelligent day by day. To start off one can think of modelling a simple system that can analyse small change is data sets and then adept to that change automatically. Or simply we can build a simple short code that can arrive to a solution without programming it to do the same.
For eg.
INPUT - Arriving to a value of 16 by any method from any two integers ;
OUTPUT - f1(X,Y) , f2(X,Y) , f2(X,Y),...
f1, f2, f3, .. are small codes that solves a problem to arrive a value of 16 from different sets of X and Ys. But each methods f1, f2, f3,.. of arriving to a value 16 are different.
The next step would be to write a software that can code an efficient algorithm from a given problem set. With these we can gradually learn how to train a computer system that reacts to different kind of situations. Finally we will be able to think of a system which can make intelligent fundamental decisions, ushering a path to our goal.
I found a blog article listing some great introductory AI courses from available coursewares like Edx, MIT courseware, Udacity, etc. here. These courses are great for beginners. If you think there are other great AI open courses available , add them in the comments. I will add them here.
While there are stripes relevant AI disciplines that normally people use to start with viz. control theory, linguistics, computer engineering, neuroscience, etc. One should directly start off with pragmatic AI project ideas without going through these academic approaches. I have my personal experience of learning NLP ( Natural Language Processing ) which concerns with the linguistic intelligence. I found that the method of learning all the tools of NLP and then going for a project is too boring. Instead one should start off with their own ideas and refer relevant books or tools to make prototypes.
Therefore my humble request to all the hackers, makers and developers is to take the next step towards AI and to all idea developers to make their ideas open source. Because I think your ideas aren't alive or worth it unless you make them happen. I just started to make all my ideas open source. :)
Presently you can share or watch DIY hardware or software projects and win prizes by taking part in online challenges. In the future users who have good knowledge in DIY electronics, hardware and software can earn money by helping other learners in our website.
Would you be kind to go and checkout livehacks.tv ? We will be giving out free goodies like T-shirts , mugs , bottles, etc. to our users. So do not forget to signup !
Thanks.
No comments
Post a Comment