“Can machines think?” – Alan Turing
AI- Artificial Intelligence- we’ve all come across this term at some point or the other. But what do these two words really entail?
In this article, I want to put aside the hype around Artificial Intelligence: self-driving cars, virtual agents, science fiction stuff (AI producing sentient machines that will take over Earth or saving earth from destruction) and the AI poster children (for instance Data and C-3PO from Star Trek). You know the stuff that we hear from the media or watching movies.
I’ve been doing some research and actually digging into what Artificial Intelligence really is beyond the noise. Through this article, I’ll share an introduction of what I’ve been reading through.
What is Intelligence?
So intelligence is something that comes in handy when we try to solve problems. It’s an combination of various skills: being creative, gathering knowledge, learning, drawing conclusions, making decisions, socializing, etc. Though there still is a lot of debate around what the exact definition of Intelligence is. I like to think of intelligence having three basic levels:
1. Gather relevant information from your environment
2. Process and Learn from that information
3. Make some meaningful decisions and perform some action
By performing these three steps on a loop we essentially “learn” to do something and perfect a skill.
Let’s connect this definition of AI to machines. An Artificial Intelligence system is essentially one that takes in information (either fed to it or it gathers information through sensors/on its own), then performs some processing, and learning from it, draws out meaningful conclusions, and maybe perform some meaningful action.
“Artificial Intelligence is the science of getting machines to Learn, Think and Make Decisions like Humans
Artificial intelligence is moving towards the development of algorithms and techniques that can simulate or even recreate the capabilities of the human mind. This will help in
- Improving Decision Making
- Increasing Accuracy
- Solving Complex Problems
- Taking Dangerous Task
- Helping in High Level of Computation
- Augmenting Humans for better performance
In a classical programming environment, we would write down a logic/code/define a set of rules and give the code some input and expect a certain output. In the Artificial Intelligence environment, we give an “algorithm” an input and “train” it with the help of data sets; essentially a data set is the a combination of the input and the output that we expect. The AI system generates the logic from these datasets through algorithms (what people refer to as “learning” from these example datasets). This is totally new paradigm of programming the system.
Through this series on Artificial Intelligence let’s explore it’s various branches: Machine Learning, Deep Learning, Computer Vision, Robotics!