The world as we know it is moving towards machines big time. But we can not fully utilize the working of any machine without a lot of human interaction. So in order to do that, we needed some kind of intelligence for the machines. Here comes the place for Artificial Intelligence. It is the concept of machines being smart to carry out numerous tasks without any human intervention.
Artificial Intelligence and Machine Learning both terms often lead to the confusion and many of us don’t exactly know the difference between them. Hence we end up using these terms interchangeably. Machine Learning is basically the learning concepts of machines through which we are going to achieve Artificial Intelligence. Deep Learning is the latest thing in Artificial Intelligence field. It is one of the ways to implement Machine Learning to achieve AI.
Most of us have seen the AI based movies with machines having its own intelligence like Terminator series, I Robot. But in real life, the AI concept was not that much optimised to handle these real-life situations and act accordingly. Mostly AI implementations were only situation based codings. Where Machine Learning was introduced to handle that much amount of data and to make the machines learn using inputs/examples to process further problems.
Artificial Intelligence
“Artificial Intelligence”, as the rough meaning suggests creating intelligence artificially, is the word that we have been repeating for more than half a century. Introduced in roughly 60’s and got everyone’s eye very soon. The goal of AI was to reduce human interaction for a machine to do it’s work properly.
AI has been implemented in several ways. It doesn’t always have to be smart implementation. Many implementations are just hardcoded functionalities which used to run according to the choices or situations. But in real time scenarios we do have a lot of variables and according to them, some action must be chosen to be executed. In those scenarios, hardcoding does not give us good results. So Machine Learning comes into the picture.
Machine Learning
Machine Learning is an approach to implement Artificial Intelligence. It is basically the study of the algorithms that make use of vast datasets to be parsed and ingested as examples and based on these examples further problems are solved.
Hence making the machine learn to solve problems by providing enough examples/inputs like a human learns something with examples and use it to solve further problems.
There are several algorithms that are used for machine learning for example:
Deep Learning
It is the newest term in the era of Machine Learning. Deep learning is a way to implement machine learning. It basically uses Artificial Neural Networks algorithm. Neural Networks are inspired by our understanding of the biology of our brains – all those interconnections between the neurons. But, unlike a biological brain where any neuron can connect to any other neuron within a certain physical distance, these artificial neural networks have discrete layers, connections, and directions of data propagation. We’ll learn more about Deep Learning in further blogs.
Conclusion
Artificial Intelligence is the broader concept which is implemented through Machine Learning (many efficient algorithms for real data). Deep learning is Neural Network based algorithm of Machine Learning.
Deep Learning has given a new level of possibilities to the AI world. Currently, Deep learning is being used in the research community and in the industry to help solve many big data problems such as computer vision, speech recognition, and natural language processing.
Still what we have today is the concept of “Narrow AI”. Narrow AI (or weak AI) symbolizes that the AI we are working on is related to some specific tasks. Like vehicle automation (Google self-drive car) or image classification or face recognition (Facebook deep learning) are some specific tasks that have been made possible with Deep Learning.
The aim of AI from the start was to create a general AI (strong AI) to achieve the functionality of human brain that was not related to a specific task but to perform all general tasks and to respond according to the situations as well i.e. mimicking a human brain processing. So we still have a long way to go.
Hope this helped 🙂
Reblogged this on Site Title.
Reblogged this on Coding, Unix & Other Hackeresque Things.