AI and computer learning is quickly gaining use, so what happens when AI becomes commonplace? Before we dive into this it is important to understand what AI can and can't do today and what aspect of it is already common. Computer learning is a subset of AI, but the two are often discussed together or interchanged. Computer learning is a method where a computer is trained on a set of data and then uses that training to learn a task. Facial feature recognition is a common computer learning task where the computer is trained to recognize the various features (eyes, lips, nose and mouth) of anyone's face.
The capability to teach machines to interpret data is the key underpinning technology that will enable more complex forms of AI that can be autonomous in their responses to input. There have been obvious failings of this technology (the unfiltered Microsoft chatbot "Tay" as a prime example), but the application of properly developed and managed artificial systems for interaction is an important step along the route to full AI. There are so many repetitive tasks involved in any scientific or research project that using robotic intelligence engines to manage and perfect the more complex and repetitive tasks would greatly increase the speed at which new breakthroughs could be uncovered. Learning from repetition, improving patterns, and developing new processes is well within reach of current AI models, and will strengthen in the coming years as advances in artificial intelligence -- specifically machine learning and neural networks -- continue.
I was at a family event recently and two guests were chatting about the Artificial Intelligence (AI) component of driverless or autonomous vehicles and more specifically, how these vehicles are currently unable to detect human movement at high speed. Form left to right, the diagram depicts a Facebook messenger bot or just as easily an independent purpose build bot placed on the homepage of a website and with the purpose of connecting with customers in a customer service capacity providing and AI-like customer service experience or response to customer enquiries. The NLP component allows the computer to interpret the vast and complicated human language, understand what's being said, process it all, reflect what is being required of it and effectively'talk back', equally like humans do. Cognitive-based systems build knowledge and learn, understand natural language, and reason and interact more naturally with human beings than traditional programmable systems.
Over the past few years, machine learning and AI have pushed forward the capacity of computers to recognize images, understand context, and make decisions. A report from IHS Technology expects that the number of AI systems in vehicles will jump from 7 million in 2015 to 122 million by 2025, bringing new opportunities to enhance the capabilities of connected cars as more data becomes available. In addition, AI will push advanced driver assistance systems (ADAS) into the mainstream. For that, they need AI, which is what enables the camera-based machine vision systems, radar-based detection units, driver condition evaluation and sensor fusion engine control units (ECU) that make autonomous vehicles work.
The state of Ohio, JobsOhio and the Ohio State University are putting $45 million into an expansion of the Transportation Research Center's (TRC) 540-acre Smart Mobility Advanced Research and Test (SMART) Center in the Columbus area. Research at TRC goes hand-in-hand with research elsewhere in Ohio, including along a Smart Mobility Corridor between the TRC and Columbus that has been primed with fiber-optic cabling and sensors that were enabled through previous funding. Lexalytics, a Boston-based text analytics software and services provider, has established what it's calling Magic Machines AI Labs at its office in Amherst to collaborate with the University of Massachusetts Amherst's Center for Data Science and Northwestern University's Medill School of Journalism, Media and Integrated Marketing Communications. With Northwestern, Lexalytics will look to identify and test real-world applications for Magic Machines AI technologies.
While I learned a great deal over the course of the semester, there was one minor point that she made to the class which stuck with me more than I expected it to at the time: before using a really fancy or sophisticated or "in-vogue" machine learning algorithm to solve your problem, try a simple Nearest Neighbor Search first. In addition, if you don't have very many points in your initial data set, the performance of this approach is questionable (though such a case in general is enough to give most machine learning researchers pause). Neural networks require a notoriously massive amount of data; this Google Neural Network paper is capable of classifying 1,000 different types of images and was trained on over a million photos. So the next time you're faced with an unknown machine learning problem, remember to give Nearest Neighbor Search a try.