With all the hype around Artificial Intelligence, Machine Learning, and Deep Learning, many of our readers (founders and students in this case) have expressed their interest to get some clarity on the three concepts. So, while there are other explanations available, in this article, we attempt to offer the simplest explanation on each of these. Most of this is based on our understanding and analysis of the concepts which are in turn based on various interviews conducted with experts over a period of time.
Artificial intelligence (AI) is a branch of computer science that aims to create intelligent machines. That’s the definition from Techopedia and we found it to be the most succinct. According to The Startup Observer, it would just be fine to call AI a “technology” if you want to. AI can be put into two categories, general and narrow.
- General AI would have all characteristics of human intelligence in a form that is indistinguishable from that of a human. Such a system has not yet been developed, and expert opinions differ on the possibility of something like that to ever exist.
- Narrow AI refers to a machine’s ability to perform specific tasks even better than humans can. Eg: IBM Watson, Facebook newsfeeds, etc.
Machine Learning (ML) is a way of achieving AI. At a basic level, ML refers to using algorithms to extract and analyse data, and then make a determination or prediction about something. Instead of writing long codes with specific instructions to accomplish a particular task, machine learning is a way of training an algorithm so that it can learn and improve based on experiences. The more it learns, the better it gets.
Deep Learning (DL) is an approach to implementing ML. Deep learning uses Artificial Neural Networks (ANNs) which are algorithms that mimic the biological structure of the brain. ANNs have discrete layers of neurons which are connected to other neurons thereby creating multiple layers. Each layer picks out a specific feature to learn, such as curves/edges in image recognition. It’s this layering that gives deep learning its name; depth is created by using multiple layers as opposed to a single layer.
In an exclusive chat with The Startup Observer in an event in Las Vegas, Celeste Fralick, Chief Data Scientist, McAfee summarised the differences among the three terms in few simple words, which we think made great sense.
Artificial Intelligence: Deals with “reasoning, logic, and problem solving”
Machine Learning: Is “automated analytics”
Deep Learning: Deals with the more “complex neurons”
Basically, Artificial Intelligence can be considered as the umbrella term that includes Machine Learning which further includes Deep Learning. The Startup Observer has created the following diagram to explain the concept.
As of April 16, 2018, LinkedIn shows 1,300+ jobs related to Artificial Intelligence in India, the highest concentration of it being in Bengaluru. While the job title mostly included the word “Artificial Intelligence”, the job descriptions often included “Machine Learning” and/or “Deep Learning” indicating that knowledge of ML and DL are necessary as part of the broader AI role.