Artificial Intelligence has become one of the most overused buzzwords in the cybersecurity field. Cybersecurity buzzwords gain more traction as the world of cybersecurity becomes more omnipresent in our cultural consciousness. Events like the Marriott Hotel Breach and the Facebook User Data scandal have created an atmosphere of paranoia. However, in the wake of an attack, people look for solutions. Artificial Intelligence has become the solution that everyone is talking about.
The concept of Artificial Intelligence has taken on a life of its own now that people outside of the IT world are starting to hear about it more and more. The overuse of Artificial Intelligence only gives people a fleeting understanding of the actual technology we are harnessing, as well as the associated risks and reward potential.
Additionally, Machine Learning and Deep Learning are also buzzwords people throw around without a second thought. Despite being subsets of Artificial Intelligence, people often use Machine Learning, Deep Learning, and Artificial Intelligence interchangeably.
What is Artificial Intelligence
Fluff posts and click-bait have slapped the word AI onto every single product that uses any sort of data processing via computing. However, AI has a very specific meaning in terms of academia, and certainly has a specific definition. AI is an intelligence demonstrated by machines. AI differs from other types of computing capabilities in that AI empowers a machine to perceive the environment it is in. Additionally, the machine will take action to adapt to the environment, using all information available to achieve whatever goal the machine has.
Some examples of Artificial Intelligence include a machines ability to understand human speech, play chess against a human, operate a car, or suggest content to you, based on your previous content viewing. Siri, Tesla, and Netflix are all large companies that use Artificial Intelligence to enhance consumer experiences.
What is Machine Learning?
Machine Learning is not separate from Artificial Intelligence. Rather, it is a subset of AI. Machine Learning is when someone programs a machine to access and manipulate data, and in turn use this data to redefine models without someone explicitly changing the programming. Algorithms and stat models are the driving forces behind Machine Learning. Machine Learning is used for common apps such as email filtering, optimization, internet fraud detection, and computer vision. Most Artificial Intelligence involves Machine Learning simply because dictating “smart” behavior requires a lot of data. Therefore, Machine Learning is the most efficient way to utilize that data.
What is Deep Learning?
Deep Learning is a broader subset of Artificial Intelligence that differs from Machine Learning in one very specific way; Machine Learning is task oriented learning, while Deep Learning is about discovering and classifying data. Deep Learning takes large amounts of structure-less data and combs through it. In searching through all this info, Deep Learning derives meaning and identifies patterns. In turn, this helps people see large scale trends or anomalies.
Some examples of Deep Learning include Google Translate, Chatbots, and adding color to black and white images.