In the past, AI was talked about for its potential in the computing market, but it was deemed not usable or applicable for a long time. Now AI is being seen in a new light. Some people, like Stephen Hawking, Elon Musk, and Bill Gates, have warned that AI may take control over humans and should not be pursued. On the other hand, Paul Allen and Jack Ma are proponents of AI.
At the recent Fujitsu Laboratories of America Technology Symposium 2015, AI was a focal point as it applied to computing. The title of the symposium was Intelligent Computing: Technology Disruptions and Social Impacts.
This blog gives an overview of the conference and what Fujitsu presented. I may write separately about some of the sessions later. Fujitsu will publish all the presentations, but its old (April 5, 2015) version of a similar presentation can be found here. The linked presentation has a ton of interesting and useful marketing data as it relates to Big Data, IoT, and other ICT concerns.
When I first saw the symposium theme, I realized that my exposure to AI has been very limited, although I have graduate degrees in computer science. But in my day, there weren’t many courses on AI. So I jumped into AI for its subfields. There are several ways to classify the subfields, but I like the one by professor John A. Bullinaria of the University of Birmingham, UK.
According to him, the subfields and their examples are as follows (those in red are covered in the symposium):
- Neural Networks – e.g., brain modeling, time series prediction, classification
- Evolutionary Computation – e.g., genetic algorithms, genetic programming
- Vision – e.g., object recognition, image understanding
- Robotics – e.g., intelligent control, autonomous exploration
- Expert Systems – e.g., decision support systems, teaching systems
- Speech Processing – e.g., speech recognition and production
- Natural Language Processing – e.g., machine translation
- Planning – e.g., scheduling, game playing
- Machine Learning – e.g., decision-tree learning, version space learning, deep learning (included by the author)
The Fujitsu executives gave their views on how AI fits into their Intelligent Computing platform, below. The three self-explanatory pictures depict three different aspects of it. The first is a logical view, the second is an architectural view, and the third is a networking view.
The logical view of Fujitsu Intelligent Computing
This picture is very easy to follow. Data generated from many sources, which is Big Data, including IoT, is aggregated and fed into an analytics stage where AI is applied. The result of that is used in many application areas, including vehicles. The Fujitsu Forum in May in Tokyo had a session on connected cars.
The architectural view of Fujitsu Intelligent Computing
One thing interesting about this slide is that it specifically shows the agriculture cloud. This is because Fujitsu is active in applying ICT to agriculture. Fujitsu is also active in the automotive cloud.
The networking view of Fujitsu Intelligent Computing
In addition to the Fujitsu presentations, I attended four panel sessions:
- AI and Intelligent Computing in the 21st Century
- Robotics and Automation
- Deep Learning and Beyond
- Future Computer Architectures for AI
I will write about some of them in future blogs.
Overall, this was a very informative symposium. I think regardless of some warnings, AI will be researched more and its results will be applied more to ICT.