2012, the Allan Turing year
Last year, in December I had the chance to participate in the World Intelligence Congress 2012 dedicated to Allan Turing. I presented a paper about my Master’s degree thesis but I also attended some presentations about research and solutions in the field of Artificial Intelligence.
The congress was held in Macao, China, between 4 and 7 December and it focused on various topics, including Web Intelligence Foundations, Web Information Retrieval and Filtering, Semantics and Ontology Engineering, Social Networks, Ubiquitous Intelligence, Intelligent e-Technology and Web Agents. These are only some of the topics discussed at the congress but you can get an idea about new trends in the field of A.I. These topics are not solutions from a distant future as one might think. We already face a lot of issues in handling large amounts of data and users. Conventional methods used in current commercial systems can become ineffective in these conditions so new techniques and methods need to be developed in order to properly scale the new generation of informational systems.
Another issue is that users are different. We know that all systems treat masses as a uniform entity, whether we are talking about political, economical or informational systems, they tend to fail because humans have different needs from one culture to another, from one age to another, and so on. Personalisation is the key to creating great systems. I think that all topics tend to converge upon this concept. Personalisation means that things are tailored for you, that you control what news and content to digest, what things are relevant for you and how the system behaves differently from one user to another to fit particular needs. In this context techniques like data mining, information retrieval from heterogeneous sources, usage of non-conventional database systems and other information storage systems like ontologies as well as the use of flexible inference engines to handle dynamically custom rules tend to become feasible solutions where conventional systems tend to fail.
Social networking is another hot topic. Social networks allow researchers to get information about users and understand the complex interaction patterns between humans. Analysing these social networks can reveal different types of groups sharing the same interests and allows researchers to identify group preferences and trends in the industry. This can have a great impact on the success of advertising campaigns or on the way we understand user needs when dealing with an informational system.
I had the chance to see various systems using mobile applications to interact with users in different fields of research. This shows that this new computing platform tends to be used more frequently not only in the industry but also in the research field, tanking advantage of these devices to develop context-aware solutions for tourism, social networking or security. By the end of the conference, a special session was held where you could vote for the most important questions in the field of A.I. I was surprised to see that all the top 10 questions were related to very abstract topics. The most interesting one was “How do we teach computers what is right and what is wrong?“. If we could teach that to computers, we could have more stable systems.
Think about the ability to understand if a program could damage our information or not (ability to spot malwares). Another question was “Can machines and humans work together?“. I was surprised to see that the speaker who answered this question was part of a research programme founded by the U.S. government. The goal was to identify tasks where machines can perform better, especially when dealing with redundant work. We face frustrated people doing redundant work every day. Increasing the automation of repetitive work could help humans focus on important tasks or it could give them more time to think and be creative.
Another important moment of the conference was the keynote of Prof. Edward Feigenbaum from Stanford University, one of the fathers of A.I. He developed one of the first expert systems and also collaborated with Nobel Prize laureates during his career. Prof. Feigenbaum seems very down-to-earth and even though he might seem old at first, his mind and his voice are those of a teenager. He had a great speech about the evolution of informatics and how early researchers were always overwhelmed about the potential of this new technology. He even stated that computers are creative and gave a lot of examples of how computers turned out to be more creative than humans. He also made a call for innovation to new researchers. They should always question paradigms and always think that rules can be improved instead of thinking how not to break them. Another thing I felt when listening Prof. Feigenbaum was his excitement about technology. I think we sometimes miss this feeling when creating software. We hide behind a lot of procedures and miss the big picture.
Overall I think the conference was a success. I had the great chance to meet open-minded people from all over the world. Even though it only lasted four days, it felt more like a month, not in a bad way though.
So what do you think, can machines and humans work together?