The original meaning of artificial intelligence
One of the major themes of the Innovative City Forum 2016 is artificial intelligence (AI). The word was first coined fifty years ago at a conference at Dartmouth College in 1956, where attempts were made to build a machine that could carry out intelligent activity like humans. Today, however, the term refers to a great many things, among them computers playing shogi and chess, assisting with housework and labor, robots used for military use, and more. While they take many forms, what most have in common is their congregation in the field of machine learning through the use of things like image searching, search engines, and unmanned driving. From the point of view of what artificial intelligence referred to, however, these are merely constituent parts of a greater whole. Another category is AGI, or artificial general intelligence. This field is closer to the original vision of "AI" -- namely, having a widely applicable skillset like that found in humans. For this year's ICF, a key focus will be discussion around the subject of AGI.
With the spread of artificial intelligence, the meaning of human existence in question
One expected subject of discussion for the event is what "work" will look like in the future. With the spread of machine learning, repetitive tasks will likely be passed off to robots and computers. The logical extension of this idea is that humans would eventually not need to work. Educational systems thus far have been devised to promote the culture of mass production following the industrial revolution, with humans performing tasks in an efficient and, so to speak, robotic fashion. In a future where AI is more advanced, however, that alone will not suffice: education will call for more creativity. At the same time, when we reach a point where can live without work, what will it mean for the significance of human existence? Work is both a means of earning money to live and a source of our various identities. In other words, we will be forced to ask ourselves what it means to be happy. This may change the role of entertainment like art and music, as well as domains like child-rearing, which are not generally considered the realm of "work" as such. The growth of AI will produce major changes to social values we now consider commonplace.
How can AI learn ethics?
As AI evolves, a key question is that of ethics. For example, take unmanned driving: is it enough for vehicles to observe the law? That is surely not enough. What choice will vehicles select when faced with immediate and present danger? When will they know when it is permissible to give the right of way? The choice of whether to prioritize law or ethics is at times an ambiguous one. For example: a car is proceeding towards a group of ten people and will be unable to stop in time even if the emergency brake is pulled. To the right is a single child. To the left is a wall. Which direction is the "correct" one? Consumers will be unlikely to want to purchase a vehicle that chooses the left and effectively destroys itself. This is the AI version of the classic "trolley problem" discussed by philosophers. Therefore, it is ever more important to discuss how to teach AI ethics. AI is traditionally "trained" through machine learning, whereby the computer is fed countless examples and cases, allowing it to develop heuristics and decision-making from among them. This is not the same as programming, so we cannot logically explain "why" a machine selected one choice. A major consideration is whether the data being fed contains bias.
One aspect of the US culture is that it espouses the rationality of law. On the other hand, Japan emphasizes ethics & morals, so we may find that Japanese television programs, with their moral lessons, would serve as good fodder for teaching AI ethics. Teaching humans and machines ethics is a very similar process.
Thus far, when discussing AI, the conversation has largely remained limited to the realm of computer science. However, elements like culture and content may become a critical piece of the puzzle. The discussion will come to encompass more creators of culture, among them legal scholars, sociologists, and psychologists.
Furthermore, we must consider what it will mean for a range of AI to coexist with humans. There is a high possibility that AI will inhabit the networks used by humans. How will each party respectively comprehend the other and interact? In other words, what will be the role of sociology and psychology in an era of artificial intelligence? Philosophers, social scientists, psychologists, and those from non-technology domains will become an increasingly important part of the dialogue. While computer science has evolved somewhat separately from these domains, going forward, we now realize anew the increased importance of interacting with the humanities.
This year's ICF may very well become a space for the discussion of these and other issues.
A kind of new metabolism for the present
The discussion on "New Metabolism," something explored in last year's ICF lectures and session, is another area we want to continue pushing forward with new lectures. In the 1960s, when Japanese architects espoused the idea of "metabolism," while the concept was large in scale, the technology of the time had not caught up, so the idea fizzled out as just a concept. In today's world of advanced technology, however, those ideas are possible. We want to bring together not only architects, but those professionals from the fields of biology, materials, and other domains to drive the discussion. Will the original idea of metabolism be restructured, or will it grow as a new movement? We don't know the outcome, but we invite you to take part.
Strengths of the ICF
There are several conferences around the world like the Innovative City Forum that explore the cities of the future. ICF's strength, however, is the way in which it focuses on not just cities and technology, but art, culture, and other domains. Frontrunners from each field congregate at ICF. When designing the future, all of these fields are indispensable to one another, and we find members of fields that would normally not all be in one place at your typical conference.
Another appeal is hosting the event in Tokyo, a place with a high level of convenient transit infrastructure. The city is also a veritable showcase for city planning, such as in the developments of the Mori Building Company and others. At the same time, it is also an experimental test bed for exciting new urban planning. For me, too, this makes the conference so interesting, as each year it brings something new that is not found at any other conference.
Joichi Ito
Joichi Ito is the director of the Massachusetts Institute of Technology (MIT) Media Lab and a co-founder and director of Digital Garage. Ito is a board member of The New York Times, the Knight Foundation, the MacArthur Foundation, and the Mozilla Foundation (the developer of the FireFox browser). Ito is a consultant to the Japanese Financial Services Agency (FSA) and a member of the FSA's Meeting of Experts on FinTech Ventures. He is also a member of the governing committee for the Ministry of Education, Culture, Sports, Science and Technology's (MEXT) COI Stream Creation of Innovation Program. Ito is chief researcher at the Keio Research Institute at SFC. In addition to being involved in the launch of PSINet Japan, Digital Garage, Infoseek Japan, and numerous other Internet companies, Ito has been an angel investor in Twitter, Wikia, Flickr, Kickstarter, Path, littleBits, Formlabs, and other promising startups. In 2008, he was selected by Business Week as one of the 25 Most Influential People on the Web. In 2011, he was selected by Foreign Policy as one of the Top 100 Global Thinkers. In 2011 and 2012, he was selected by Nikkei Business as one of the 100 most influential people for the future of Japan. In 2011, he received the Lifetime Achievement Award from the Oxford Internet Institute. Ito received the degree of Doctor of Literature, honoris causa, from The New School in 2013. In 2014, Ito was inducted into the SXSW Interactive Festival Hall of Fame. He received the Golden Plate Award in 2014 from the Academy of Achievement. In 2015, he received the degree of Doctor of Humane Letters, honoris causa, from Tufts University. On July 1, 2016, he was appointed Professor of the Practice in Media Arts and Sciences at MIT.