Future trends in AI Research (Sparrho AI Feature: Part 2 of 2)

Last week we spoke about the recent milestones in AI research. Now we try to answer the next question — what next? Where will AI research go and where will it take us?

To start off the discussion, last week we spoke about how IBM created an AI that is able to beat people in Jeopardy! IBM is investing $100 million in Watson — however. the question to ask is, surely, what is the point of this all this [ 2 ] ? The two main areas that IBM plan to use Watson for is research analytics and services. They plan on finding a way to harness the ability to discover information through big data and help researchers, initially targeting pharmaceutical research to reduce research time from months to days and days to hours.

While Google has developed AlphaGo, which is able to beat a professional player in Go [ 3 ]. According to the head researcher, one of the major goals of AlphaGo is to create an AI that is able to master complex disease analysis and climate change modelling [ 4 ]. Even Facebook is ramping up their AI research with FAIR (Facebook Artificial Intelligence Research) [ 5 ]. They have been working on objection recognition, natural language, and predictive learning. Their goal is to make an AI capable of seeing and understanding.

Technology and research-wise, the next big area to watch is deep learning networks, also known as convolutional neural networks [ 6 7 ]. This is a very interesting area of research as it combines explosion in computing power and networks by using cloud computing, and the heightened speed of AI development.

This concludes our two part mini series on AI research. Tell us what you think! Where do you think AI research will take us? Do you agree with our predictions?