The Surge of Deep Learning

By CIOReview | Monday, April 9, 2018

Many would say that the moment AlphaGo, the artificial intelligence (AI) powered machine, beat the world champion at the game of GO, the power of AI was showcased to the world in all its glory. GO is not just another board game that an average person can play, but a detailed strategy oriented mind-bender that even the greatest of minds fail to master. The developers of AlphaGo emphasize that such AI (general purpose AI to be specific) could play a vital role in medical and clinical sciences in understanding complexities of the human body. Today, artificial intelligence is used in various walks of life, if not to its full potential, but enough to automate a significant amount of redundant and manual processes.

The debate of AI is the next big thing that exists solely because AI can only get better in time. Deep learning and machine learning capabilities allow data-driven technologies to rectify the mistakes previously made, and improve upon existing datasets for enhanced performances. Consider natural language processing for example. Google translate, one of the very efficient natural language processing applications learns user’s voice, tone, and modulation to understand regional dialects, accents, and pronunciations. With a year of usage, the application would have even learnt various keywords that the user would speak or type, enabling predictive text and speech for faster inputs. At one point in time, the same user can directly dictate actions to his phone using mobile assistants instead of going through various menus to perform those actions manually. These advancements only speak about one thing—the growth of AI. Whether it becomes beneficial in the long run or eliminates human work completely is a different debate altogether.