The mind is a machine. Though not much of a stretch these days, prior to 1930 this way of thinking was non-existent according to Margaret A. Boden, author of “Mind as Machine: A History of Cognitive Science.” Cognitive psychology, the study of the mind as an information processor, gained prominence in the 1950s as computer-powered machines became more advanced. Cognitive science exercises concepts from artificial intelligence (AI) and control theory to understand the manual and autonomic functions of the brain. Fast forward over a half century later, and a subfield of artificial intelligence — machine learning — is the centerpiece of innovation for major tech conglomerates.
AI Conquers Go
Machine learning techniques and algorithms are being developed constantly for a variety of uses. Developers at the company DeepMind (acquired by Google in 2014) took machine learning to the popular Chinese game “Go,” in which players make long series of strategic decisions using multiple factors in human logic. In March of 2016, a top-ranked “Go” player, Lee Se-dol, lost three straight matches before winning one against the AI machine learning competitor, “AlphaGo,” developed by DeepMind.
Following the initial matchup, humans had a shot at redemption this past May when the world’s top ranked “Go” player, Ke Gie, squared up against “AlphaGo” for a best of 3 set. “AlphaGo” swept Ke Gie in two straight games, meaning artificial intelligence and machine learning capability has surpassed human ability in the world of “Go.” This is just the brink of where this computer logic can be applied.
Machine Learning Transcending Medicine
Using a similar algorithm used with “AlphaGo,” Mark Waller and his team from Shanghai University in China, are developing a streamlined computerized system for creating molecules. This process has been a heavy time-suck for chemists over the entire history of its existence, requiring trial and error under numerous testing conditions. The most recent evolution of digital molecule creation “Syntaurus” analyzes over 12 million known reactions before suggesting the best pathway for synthesis. With this technology, successful pre-discovered syntheses were mapped out within seconds, matching the exact compositions of verified literature. The form of machine learning used in both “AlphaGo” and “Syntaurus” is known as Monte Carlo Tree Search (MCTS). Although it has been demonstrated to work in multiple cases, there are still other variables such as the relative spatial arrangement of atoms and natural products that have not been accounted for in the algorithm as of yet, but it’s certainly an exciting step in the future of medicine.
Mobilized Machine Learning
Anytime new technology is developed in this day and age, it should be expected to extend to our mobile devices and machine learning capabilities are no exception. Mobilized machine learning can have a profound effect on our everyday lives once fully realized. Qualcomm is one of the leading innovators in mobile machine learning, and their Snapdragon mobile platform is a shining example. The technology memorizes user tendencies and in return constantly adapts to better focus on the needs of the user. This leads to increased battery life and efficiency, higher quality connections at all times, streamlining of manual processes and adaptable security to ensure no other user can falsify identity or access critical information. Beyond enhanced practical usage, Snapdragon also offers the mobile machine learning luxury of an intuitive camera. Objects and surroundings in pictures/videos will be recognized by your device for improved compositional value and post-production editing.
Machines have minds of their own, but prior to the 1950s, those minds couldn’t conceivably think independently. The entire script has flipped with the advancement of artificial intelligence. Machines help us think and operate, and can now communicate back with advanced analytics while mastering our thinking patterns.
Conspiracy theorists and some futurists fear the day when machines will outsmart people to the point of frightening consequences. That’s a separate debate. For now, let’s rejoice in the beautiful innovations machine learning capabilities have provided and will provide in the near future.