AI is back in popularity after a long winter (Prof Zadeh), IBM researchers believe that AI winter was over in 2011. The popularity of AI is resulted from the success of marketing based or customer based AI (Stuart Rubin). Smarthpone is an intelligent product for marketing and customer. However, the AI reseachers have been working on AI issues since 1955 or 1956 based on their curiority on intelligence in terms of science and technology, although they sometimes look at marketing and customers of AI. Human Intelligence is still a lasting issue for us to study in terms of science, technology, engineering, management, systems and tools. Fuzziness in human intelligence is also an attracting topic, because it is a strategic asset of human intelligence. Thereforee, AI will attract more ressearchers to undertake research and development on computerization of intelligence or humanization of machine. Fuzzy logic will still play an important role in developing AI.
The current problem in AI and fuzzy logic is: Can we list 10 hard problems for AI and fuzzy logic for researchers and developers. These 10 hard problems are so important adn hard that they cannot be solved using AI and Fuzzy logic within 10 years. These 10 hard problems can also attract researchers and developers to selflessly contribute to solve them day and night.
I do not believe that deep learning (Li Chen), smartphone, neural networks (Paul), image recognition, speech recognition and natural language processing are one of the 10 hard problems, because all these are already traditional products or technologies. They cannot bring about revolutionary development for AI general and in themselves in specific.
I still believe that deception is a kind of human intelligence. Without deception, there is no human intelligence. We love intelligence, intelligent systems and smart products (e.g. smartphone). We hate deception and fraud. This is reality. Therefore, AI have been customer-based since 1956. We can blind ourselves from deception as an intelligence, fraud as an intelligence. Maybe science of deception, computing of deception is one of the 10 hard problems for AI for researchers and developers. A fuzzy logic approach to deception is a good topic as the starting part in this direction.
Thank you for your attention.
Dr. Zhaohao Sun, PhD, MACS (Snr) CP, MIEEE, MAIS
Associate Professor & HoD (Chair)
Director of Research Centre of Big Data Analytics and Intelligent Systems (BAIS)
Chairman of Postgraduate, Research and Publication
Department of Business Studies
PNG University of Technology, Private Mail Bag Lae 411, Morobe, PNG
Phone: +675 473 4406 (O), Fax: +675 475 7667
Categorization and planning in context of Laws of robotics, e.g. robot needs to think a lot about current situation and thinks in which category his actions falls before doing anything.
I think in this question there are two different levels of analysis. AI is a field of research, which can include Fuzzy Set Theory/Logic, as a tool for representing uncertainty. But even in this function is not the only one. There are several formalisms that can deal with uncertainty.
On the other hand, Fuzzy Set Theory has its own independent existence, as a way to represent membership relations, different from that traditional of Set Theory.
Regarding AI current problems, I think semantization can be one of them; It could be understood as how machines understand the meaning of human expressions.
- Badges
- Science topic