General AI
This page is under construction.
'Intelligence' is a trait that is normally associated with only humans (or animals
to some extent). AI or Artificial Intelligence is the science concerned with
intelligent behaviour in machines. The goal of AI can be broadly defined as design
of machines that do things as well as humans. AI concerns many activities
like game playing, prediction, image processing etc. AI in computer science has
evolved as algorithms that learn from their previous experiences, do intelligent
searches (rather than use brute force methods) etc.
Some such algorithms are Neural Networks, Genetic Algorithms, Fuzzy Logic etc.
I plan to discuss some of my favourite AI methods on this page.
Some say that it is not possible for machines to think while some fear that
if machines do learn to think and attain what we call conciousness, that might
cause problems for us humans. Right now, even them most basic brain
(maybe that of a cockroach) is much more complex than the most complex computer.
Maybe in the future we will have machines that do think.
A school of thought says that maybe our brain (yours and mine) is just a computer,
a really really powerful compter. It has the traits of being extremely high speed,
highly distributed and massively parallel. It is so powerful in processing its
inputs and responding to them that it appears to us that we 'think'. If
this is really the case (and we dont know as yet), then maybe in the future
(or is it near future) we will have machines that
think and behave like humans. Only time will tell. There
are many systems that come under the realm of AI. Some of these are :
Expert Systems
- These are decision making systems that make decisions about real world scenarios
which would otherwise require an expert. These are not truly intelligent but
many powerful expert systems exist and are in use today.
Agents
- An agent is basically an entity (normally software) that acts on the behalf
of some other entity (could be you) that exhibits some of the traits of the
latter.
Robotics
- AI finds major applications in the field of robotics. It wont be long before we
have robots who can see, hear and feel (just like us) and act on these stimuli.
Neural Networks
- Neural networks is basically a network of artificial neurons (software replica
of brain cells) that can learn from experience (called training of the network).
They could form the underlying brain of your intelligent software as they can be
designed to learn and adopt over time. [This topic is my personal favourite]
Genetic Algorithms
- Another interesting topic. These algorithms are mainly used to perform intelligent
search and equation solving. Such a search has many generations. Each generation
has parents which mate to generate offsprings and only the healthy ones survive.
The healthy ones are the ones most probable of taking us to the right solution.
This ultimately leads us to the result.
Fuzzy Logic
- In the real world nothing is black or white. Right or wrong. Shades of grey
do exist. Digital computers, think it terms of zeros and ones and and the intermediate
state does not exist. Fuzzy logic is concerned with designing systems
which have intermediate states also, as against the standard yes and no. States
like 'maybe'.
Cellular Automa
- This basically started off as Conway's 'Game of Life'. Which was basically
a 2D system in which colonies either multiplied or became extinct depending on
the population density. They had only two states (alive or dead). Now it has
many variations, with many states and many different types of cells.
Parth Malwankar
Contact :
[email protected]