[Feature: AI Society and Public Space] Two Implications of "Publicness"

2019/02/05

Consider, for example, the claim that a city square is a public space and must be open to everyone. Some may recall that in Western cities, such spaces are often provided facing churches or city halls, and are used by various people as Sunday markets, stages for street performances, or places for political appeals. A square is a place that can be used freely in this way, and because of this, everyone can enjoy various conveniences. Now, does that "freedom" include the freedom to drive stakes, wrap ropes around it, claim it as one's own, and occupy it?

Everyone would likely answer "no." Being open to all means that the possibility of access must be guaranteed (even if potentially) for everyone, and it is not permissible for one person to restrict the freedom of others by dominating the space. The freedom guaranteed here is freedom only insofar as it does not infringe upon the similar freedom of others.

A space with a public character thus means, on one hand, that a guarantee is given to me that I can use it at any time, but on the other hand, all others are also guaranteed similar usability, and I must tolerate the use by others (and thus the fact that I am excluded from the space currently being used by others and cannot use it simultaneously). The idea of equality behind publicness—the principle that everyone is one person and no more than one person, and cannot stand in a superior position to others—brings about this duality of freedom and tolerance.

The question then is whether the fact that I must tolerate use by others, as just stated, includes the fact that I myself am seen by others and exposed to their gaze. In Japan, for example, it is considered that people inevitably have the possibility of being seen by someone else on public roads or similar public spaces, and generally, "being seen" itself must be tolerated. For this reason, it was judged lawful for the police to film a suspect on public roads and inside pachinko parlors to determine identity with a person caught on security cameras (Supreme Court Decision, April 15, 2008). Even in a case where a violation of privacy was claimed because laundry hanging on a balcony was captured when Google Street View photographed scenery from a public road (Fukuoka High Court Judgment, July 13, 2012), illegality was ultimately denied on the grounds that it fell within a certain limit of tolerance.

In contrast, the attitude of EU countries is generally said to differ greatly from this. For example, Swiss privacy authorities once expressed the view that it is impermissible for dash cams to film the surroundings from inside a car, as it is impossible to obtain prior consent from all individuals who might be captured. In countries like France, warnings have also been issued that even when installing a security camera at one's front door, it is illegal to set it to film the sidewalk ahead beyond the premises. Behind this is the idea that since roads are precisely public spaces and must be open to all people, it is also impermissible to exclude people who, for some reason, do not want to be filmed or seen.

Some may doubt whether there is such a thing as a legitimate reason for not wanting to be seen by others. For example, if someone is hurrying down the street to commit a crime right now, they would want to avoid public notice, but we would rather say it is legitimate to monitor them against their will. However, is it inappropriate for an actor, who is so popular that they are always surrounded by fans if noticed, to want to spend a rare holiday quietly? It is also a reality that people with large scars or abnormalities in their appearance, such as on their face, are looked at with curiosity by those around them (or worry and suffer thinking that they will be). Even for those who argue that for actors it is a "fame tax" or an unavoidable part of a self-chosen career, they surely cannot deny the existence of people who "do not want to be seen" for reasons where self-responsibility cannot be questioned, as in the latter case.

In other words, we can consider that a conflict between the two implications of publicness pointed out earlier is occurring here. When the desire for me to walk freely in a public space contradicts the fact that others who can enjoy that freedom just as I do exist in that space and I am seen by those others, which one should we prioritize and for what reason?

Consider, for example, the question of whether a person walking in a square is seen by others, or whether a conversation on a public phone is heard by those around them. Of course, with a public phone, a Western-style response of physically blocking sound by putting it in a box and dividing the space would be possible. However, it would be simply impossible to divide a square into countless paths where gazes do not pass through each other to protect the freedom of movement of those who do not want to be seen, or even if it were possible, it would result in creating something that can no longer be called a square.

To begin with, the human eyes and ears at issue are things that cannot be closed on one hand—things that automatically see or hear unless very strong coercion is applied—and it must be said that the possibility of success is low even if one attempts a physically complete response to them. On the other hand, they were imperfect existences where even if something should physically be heard, it is unknown whether they are actually listening (paying attention and recognizing it), or whether they will remember it even if they heard it. That is why, in comparative cultural studies and the like, it has been emphasized that a Japanese-style solution of handling things through manners, such as "pretending not to hear" conversation content that should be audible from a phone placed in an open space, was established.

Considering that behavioral patterns such as "pretending not to see" each other in places where it might be slightly awkward to be known to be frequenting have likely developed regardless of East or West, one could think that even in the West, people have trusted this human nature somewhere. By trusting in human imperfection and social customs, public space was positioned as a place where no one is seen or heard while being open to everyone, and conflicts regarding publicness were avoided.

However, in recent years, the "machine eye" newly born from the development of information technology is different; it possesses the perfection of recording various things as set and remembering them unless deleted. With the generalization of surveillance using electronic technology and the emergence of a situation where information collected there is automatically and rapidly analyzed by AI, it can be said that the question of to whom and to what extent public space should be open has become something that should be seriously considered.

And this point will be questioned more seriously on the internet, which has become an important part of our daily lives as a route for exchanging information, even more than in the electronic surveillance deployed on the streets where we actually come and go. On one hand, we naturally have the sense—regardless of whether we call it information privacy or the right to control one's own information—that information such as what books I bought at an online bookstore, who I talked to on SNS, or which pages of which sites I viewed is "mine" and we do not want it to be known or used by someone else without permission.

However, in these actions, there is always a counterparty first; bookstores, conversation partners, and site creators inevitably receive that information, and in certain cases, they are expected to respond appropriately based on it (an online bookstore that does not recognize an order and thus does not send a book would have no reason to exist). In the case of SNS, that conversation might be placed in a state where it can also be seen by users of the same service, and even in an online bookstore, my actions might be influencing someone else's actions somewhere, albeit in an indirect form such as affecting the creation of "recommended" lists ("Customers who bought this book also bought..."). In this sense, the internet is a public space where various information from various people comes and goes, and just like physical space, the relationship between freedom and tolerance in that place should become an issue.

For example, the reason the "recommendation" function mentioned just now can continue to be useful for us is likely because many other users are seriously choosing products and buying what they truly need. As actually became a problem at one time on Amazon.com, if people are allowed to freely write comments and ratings regardless of whether they purchased the book, it leads to a flood of criticism against the author's personality or actions rather than the book's content, or conversely, a line of contentless and uniform rave comments by entities that seem to be hired for promotion. It is easy to understand that such "unreliable evaluations" greatly damage the convenience of the service and the value of the evaluations if one recalls general-participation gourmet sites and the like. Fake news, which has been a frequent topic recently, can also be thought of as a problem created by such treacherous users of the internet. If we do not behave ourselves so as to exclude the treacherous actions at issue here and not restrict the appropriate enjoyment of the service by others, the freedom of the public space we ourselves enjoy will perish.

And it can also be said that what is necessary for that is the possibility of tracking and evaluation. For example, even in ancient societies where the state and laws as we know them did not exist, economic activities must have been carried out to some extent. Even in situations where one cannot force a contract to be kept through a trial, people make promises to each other and can maintain economic relationships such as buying, selling, and leasing by voluntarily keeping them. Trust and evaluation are what people called libertarians, who take a negative stance toward the state in general, have cited as important factors for this. A person who does not keep a promise they made will be evaluated by those around them as having the possibility of doing the same thing next time, and will no longer be able to have new promises made with them. A rational person who can understand that this leads to their own disadvantage should try to avoid actions that lower their evaluation from surrounding people and lose trust as much as possible. In this way, people try to keep promises autonomously and voluntarily, and we can also make promises trusting that other people will do so.

But as is immediately apparent, for this process to function, it is necessary to understand who is making the promise and what the result was from the perspective of surrounding people, and for the resulting evaluation to be accumulated. In an ancient communal village, those conditions might have been met, and it might have been possible within a very limited social group like the great merchants engaged in medieval Mediterranean trade. However, on many SNS, one can create countless accounts that do not necessarily correspond to the actual me—an online "me"—and if some misconduct is discovered, one can discard it and be reborn as a new "me." In a situation where the "me" as the subject of tracked behavior and accumulated evaluation is highly fluid, a strategy of aiming for short-term gains through treacherous actions becomes more advantageous than building up good evaluations to win the trust of those around.

Similarly, if the party that broke a contract could freely demand that the information be deleted, the mechanism of trust would again cease to function. The certain degree of anonymity realized by the internet and the portability between services and accounts provided us with a free space liberated from the human relationships and pressures that arise in real society, creating great convenience. At the same time, however, by making the accumulation of evaluation difficult, it also caused the loss of the possibility for a spontaneous order to develop and be maintained.

Under these circumstances, if we wish to maintain the convenience brought by the free distribution and use of information through the internet, we will face the following choice. If we want to avoid the order of information being disrupted by the existence of treacherous users, individual users—that is, each and every one of us using the internet—must avoid such acts and appropriately transmit and allow the use of information so as not to restrict the convenience of others (at least intentionally). If appealing to the awareness of individual users does not sufficiently deter problematic behavior, the possibility of tracking and evaluation must be guaranteed by some systemic response. As introduced earlier, the system of allowing only users who actually purchased to post reviews and ratings can be understood as an example of this. Conversely, it should also mean that the idea of treating all information about an individual as if it were his exclusive property and allowing free control by his will should be suppressed.

Let me repeat the conclusion. In order for a public space to bring its convenience to everyone as an "open place" without losing its publicness, the restriction that exclusive occupation by a specific person must be excluded is indispensable. To enable personal information accumulated and distributed in today's information society, centered on the internet, to generate benefits for everyone, treacherous users who try to monopolize benefits for themselves by distorting information should be excluded, and a certain gaze of suspicion might also need to be directed toward attitudes such as enjoying the convenience generated by information provided by others while not giving up "one's own information."

Publicness inherently includes the premise of not restricting others and the limit of having to tolerate others enjoying the same freedom. If we are to enjoy the benefits of AI technology while preventing its harmful effects, we will need to seriously consider institutional mechanisms that can find and maintain an appropriate balance between freedom and tolerance.