Artificial Intelligence, Machine Learning, Deep Learning. These terms are all common in tech circles. They evoke visions of self-aware robots and sentient computers. But fear not, sentient AI continues to remain in the realm of science fiction, at least for now.
In Computer Science, Artificial intelligence is a catch-all term for statistical models that can be used to make deductions based on provided data. Machine learning and deep learning are sub-fields of AI that methods for creating and training these models. Models are trained using data that has known conclusions. Once a model has been trained, it can then be used to analyze unknown data and provide a conclusion with some degree of confidence. Algorithms are developed and trained based on the problem that needs to be solved. There are many different disciplines within AI including Natural language processing, computer vision, and probabilistic modeling.
Natural Language Processing is a field dedicated to understanding the written and spoken patterns of human speech. Siri, Alexa, and Bixby all use natural language processing to interpret voice commands and to take action on the command. On the surface, it would seem like giving an algorithm a dictionary of words and a set of grammatical rules would be sufficient. The challenge lies in interpreting slang, colloquialisms, speech patterns, dialects and intonation.
Computer Vision is used to infer information from an image. Facebook uses computer vision algorithms to recognize faces in images. Tesla uses computer vision in their autopilot system to analyze the road and surroundings to make sure the vehicle stays in its lane and avoids collisions with other vehicles or objects. Galaxy Zoo uses computer vision to classify galaxies using images from space and radio telescopes.
Probabilistic Models are useful for determining the chance of a given outcome on unknown data using observed trends. Google, Amazon, and Netflix use probabilistic models to predict a user’s behavior. Google’s search engine will suggest alternative searches by analyzing what you typed in and what other users have searched for. Amazon shows you products you might be interested based on what you’ve purchased in the past and similar purchases by other people. Netflix suggests shows and movies using information on what you’ve watched before.
AI has far reaching applications from biology and medicine to business and finance. What can AI do for you? Let’s say you have a customer relationship management (CRM) application. You use it to track the sales process from new leads to closed deals. You could use AI to find trends in your data that can be used to improve the sales process. For example, your AI algorithm has recognized that deals are rarely signed in the months of November and December but deals initiated in January are almost always signed. Knowing this, your company could create a targeted marketing campaign in January to help close more deals.
The last major revolution in computing allowed companies to start amassing large quantities of data. The question then was "what do we do with all this data?" AI is the new revolution. Where AI was once exclusive to researchers, mathematicians, and computer scientists, it is now becoming available to the general public. There are a wide variety of applications that give businesses the ability to leverage AI without having to write code. Azure Cognitive Services (ACS), part of the Microsoft stack, is one such service. (ACS) provides several components that can be used to analyze your business data. ACS integrates seamlessly with other applications in Microsoft 365 like PowerApps and Flow. You could analyze and tag images, read the text from a business card and add it to Dynamics, or create your own custom model to analyze your data. All this is at your fingertips with ACS.
In the next blog post, we will go into detail on what Azure Cognitive Services is and what it can do for you.
If this blog piqued your interest and you want to know more how AI and Machine Learning can help your organization, please Contact Us!
5 Reasons Most Technology Projects Fail