What is machine intelligence?
For those of you that have an interest in the world of technology, whether personal or professional, you have almost certainly encountered the term “machine intelligence” in recent years. Machine intelligence is not a new concept. In 1959, Arthur Samuel defined machine intelligence as “the ability of a machine to generalize knowledge from data without being explicitly programmed.” (1) In other words, machine intelligence refers to the ability of a system to recognize patterns and to make predictions based on these patterns without human intervention. The computer “learns” by continuously gathering data to improve the accuracy of its predictions. Since the term was coined, programs have been created that have simplified the use of machine intelligence, allowing machine intelligence tools to be widely applied to anything to do with the processing of data.
Many of us don’t realize that machine intelligence is a big part of our everyday lives. When ATMs are reading the written numbers on the cheques we deposit, Facebook is automatically tagging and identifying our friends and Siri is giving us those directions we asked for, we are seeing machine intelligence in action!
The potential of machine intelligence expands beyond the everyday. IBM has created a robot named IBM Watson that is making groundbreaking progress in artificial intelligence. In 2011, IBM Watson competed on Jeopardy against former winners Brad Rutter and Ken Jennings. IBM Watson had access to four terabytes of storage but no internet access — and won! (2) In other words, IBM Watson was able to identify the question being asked, predict what the best answer by running probabilities against storage data, and provide a verbal answer faster and with a greater degree of accuracy than the experts themselves.
Today, IBM Watson is learning how to identify objects using visual recognition, a machine intelligence application. Check out this video of IBM Watson using visual recognition to identify a hand of cards. (3)
IBM is not the only company making breakthroughs in artificial intelligence. Google Glass has teamed up with Dapper Vision to design a product called OpenShades. (4) OpenShades uses speech recognition and text-to-speech machine intelligence applications to answer questions about the physical environment. For example, when shopping at the grocery store you can take a picture of the dairy section and ask, “where is the 2% milk?”. OpenShades will send the photo to Amazon’s Mechanical Turk, a crowdsourcing Internet marketplace, and Twitter to ask the public for help and the answer is read out loud to the user. OpenShades has tested the base models on users to demonstrate how this technology can benefit people who are visually impaired.
Imagine combining the OpenShades product with IBM Watson’s visual recognition capacities to create a product that could give visually impaired individuals real-time answers about their physical environment.
We have only started to scratch the surface when it comes to the potential of machine intelligence. It may be that the only limiting factor to find new ways of applying machine intelligence techniques is our imagination!