Smart Cities Hijacked by Technology | Book Review

Appropriate application of technology to regain control of the future of our cities

The Smart Enough City: Putting Technology in Its Place to Reclaim Our Urban Future

For more articles, welcome to follow【 SeiferTalks 】

The author of "Smart City Kidnapped by Technology" is Ben Green. According to Zhe Shukou's introduction, he was still a doctoral candidate in the Department of Applied Mathematics at Harvard University when the book was published, and served as a director of the Boston Innovation and Technology Department from 2016 to 2017. data scientist. The Chinese translator is Liao Tingyun; published by Pedestrian Culture Lab in August 2020.

Smart City, before and after the appearance of this term, was full of ideas or policies that wanted to apply various (latest) technologies, networks, big data, and artificial intelligence to urban life (according to the author's point of view, it can be seen as early as 1930). years later). They have been inundated with government advertisements, election propaganda, and even articles large and small whether or not they may be "matched" by the companies concerned. It's "horrifying" that we can't pinpoint when it started or where it happened, but we all know that we hear or see these kinds of claims almost often.

This is the main argument of the book "Smart City Kidnapped by Technology" : the use of technology to govern municipal administration has probably gotten rid of the supervision of democratic mechanisms and responsibility politics, and the surveillance even affects everyone unknowingly .

The term "kidnapping" is undoubtedly used by the Chinese publishing house for the sake of alarmism, but to some extent: due to the knowledge gap within the complex technology (for example: hardly any officials or citizens really understand the algorithm inside the machine) how it works), city officials and the public often only accept or accept it with a half-knowing. Once adopted, it is difficult to supervise or be held accountable, because technical aspects such as algorithms involve the business secrets of enterprises, and the city government can even use "smart information" and other reasons to set a high threshold for people to explore and review them. The whole city seems to be bound by countless signals, cameras, and lines.

The author of this book has a solid background in science—especially in the field of "mathematics"—theoretical and practical experience, and what is even more rare is that in his involvement in such a large and international urban civic business in Manhattan, he did not settle for a single department. In contrast to the stable "little days" in the bureaucratic system, it is alert that in the face of the real world full of "living people", pure natural science and electronic computers alone are absolutely not enough to take care of the extremely complex diversity and diversity of human society. Derivative issues. To this end, Ben Green fully extends the research and exploration of knowledge to the fields of political science, sociology, economics, etc., in order to more comprehensively understand how to avoid making biased and incomplete decisions based on technology alone .

In my opinion, this is a good book that every civil servant in Taiwan (including public opinion representatives) should have in hand , and even the mayor, bureau chief, or division chief should hold a reading meeting every day to discuss the content with the staff.

Starting from the cases that the author personally participated in and collected, "Smart City Kidnapped by Technology" discusses the "privacy", "criminal issues", and "democratic participation/accountability" of the people caused by excessive advocacy of "technology governance" one by one. Infringement statement. Although most of the cities listed are in Canada and the United States, many of these problems are also found in Taiwan—especially in big cities like Liudu, which are competing for development.

The book is full of clear examples, supplemented by the author's caring analysis from both science and technology, politics and humanities, and makes an impressive review and warning in just 286 pages (and attaches a lot of references and references quite in line with his academic status. citations). It's quite challenging to write a review of this type of work, so I'm not going to do too much of a roundup, but instead echo and reflect on Ben Green's five suggestions at the end of the article.

How can we avoid the bias of "smart city" and advocating technological governance from causing harm to the entire city and the people living in it? The author summarizes five principles that run through the book:

1. Deal with complex issues rather than superficial ones

Here are simply two examples in the book to make the reader understand.

MIT (Massachusetts Institute of Technology, Massachusetts Institute of Technology) proposed a "smart intersection" traffic plan in 2016, which can allow "Automating Vehicles" to travel freely in the city, but it has become the most classic Absurd case of:

Once this new technology is deployed, traffic jams will become the ashes of history, and these simulations of future streets seem to herald the dawn of a new era where advanced technology will effectively alleviate long-term urban traffic problems.

However, the mathematical models and simulations devised by MIT still fall short: crowds. Their city streets are devoid of life other than traffic , and this absence is particularly noteworthy because the intersection at the heart of the MIT model is actually the busiest pedestrian and transportation hub in downtown Boston. It is also one of the most walkable places in the United States.

Another example is the use of big data to analyze, calibrate and even "predict" which areas are hot spots for crime in a city, as a kind of "Predictive Policing". However, human rights groups and even high-level police officers have found that while the use of such technology can reveal where crime rates are high, it also allows police officers to take measures that are more likely to violate human rights when they are on duty. When officers enter the alert range at the time the system is alerted, they usually become "extremely alert"; this allows officers to "treat everyone they see as possible criminals." In addition, because the data of machine learning is a record of the past, and the long-term racial discrimination in the United States has resulted in a high rate of arrests of people of color and immigrants, the recommendations made by computer analysis are full of "neutral bias".

To take a deeper look: the arrests or high crime rates in some areas are actually due to the shortage of “medical resources” or the distance, which makes people with physical and mental illnesses lack the opportunity to receive medical treatment or continue to see a doctor on a regular basis. In 2015, Lynn Overmann, then a senior adviser to the head of the US Department of Science and Technology, said:

America's largest mental health facility is often our local prison

In other words, when faced with problems such as traffic congestion and reducing crime rates, rushing to adopt big data and artificial intelligence on the surface is often counterproductive. Because "machine learning" is "garbage in, garbage out" (garbage in; Garbage out): the vast majority of data fed into high-performance computers and "feeds" to algorithms is accumulated in the past, if those records themselves are Full of prejudice, the final transmission from the electronic system will also be full of discriminatory references . On the other hand, the seemingly causal relationship may not be as seen, perhaps a "spurious relationship" in statistics. One example is the previous link between people with mental illness and areas with high crime rates. If you only look at it from the surface: some areas are prone to crime simply because there are many homeless people, so driving homeless people or strengthening patrols has become an important task. But under the surface of the poor homeless is actually a mental illness caused by economic hardship. The average person doesn't attack for no reason unless he can't control himself or doesn't know what he's doing at all.

Driving more roads when it is easy to get stuck in traffic, sending people to prison (or clamoring for the death penalty) if there is a problem, don't these sound like the current situation in Taiwan? Is it really better after doing this? Many examples in the book give the exact negative answer.

2. Applying technology to address social needs and drive policy, rather than making goals fit the function of technology

We originally wanted to improve the efficiency of public facility repairs through "people's supervision" and allow ordinary people to report to deter and reduce vehicle violations. "It's unbelievable. This is an example of "pure efficiency" without adequate consideration of negative effects.

The book uses an example of a "311 app" in the United States to point out that increasing the speed at which people transmit the location of potholes to be filled will not increase "true civic engagement." In fact, it reduces the willingness to engage in democratic politics, because people think it's okay to "teleport the problem"; deeper discussions and accountability are skipped. In addition, those who can use the APP to report or raise issues at any time are often young people, people from the middle class and above: they know how to use it and have a wireless network at all times. What about people who don't have a mobile phone, don't know how to use it, or even have the money to pay for the internet? Another kind of discrimination and subclassification arises spontaneously.

Participatory budgeting is the author's solution. Encouraging and guiding the public to actively participate in the discussion and deliberation of municipal issues is the real civic participation. Similar arguments have actually been made in Taiwan for a long time, but government agencies often regard people with many problems as "tricksters" looking for trouble, and are afraid that such participation will hinder the governance and performance of public servants. Practice it well. The "Youth Council" or "Youth Consultation" implemented in some counties and cities is even more useless with embroidered pillows. In the early years, government units would prevaricate with reasons such as "insufficient internet" and "insufficient technology", alleging that a large number of people will be online at the same time or too many users in the discussion area will cause the system to be overloaded. The significant reduction in data processing costs has not yet been implemented.

3. Prioritize policy and program reforms rather than the use of new technologies

This can be regarded as an extension of the second principle. Solving problems should not be about finding a technological tool and then seeing how to use it, but about finding out the real cause-and-effect relationship first.

Take mental illness and crime rates as an example. When the police rely solely on crime hotspots and related computer predictions, "catching people faster and sending them to jail" seems to be the most innovative approach. However, when the mentally ill, low-income households, or single-parent families with financial difficulties do not have access to good medical care and education, or when social resources continue to be unequally distributed and fail to appear in areas of extreme need, detention centers and prisons will continue to suffer. Overcrowded; people released from prison may be in worse condition. I only see and only want to use big data, artificial intelligence, and algorithms to solve human problems, and usually only repeat the mistake of "solving the symptoms but not the root cause".

4. Ensuring that the design and application of technology safeguards and promotes democratic values

Whether in the United States or Taiwan, "democratic values" are an important foundation for national institutions. Among them, "citizen participation" and "responsibility politics" are the most critical.

The indifference to politics, elections, and policies gives the bad elements in the bureaucracy the opportunity to attack and abuse power and corruption. Putting "private business interests" or "pursuit of efficiency" ahead of adequate oversight would undermine people's rights and interests, including (but not limited to) the right to privacy, suffrage, the right to request information, and even the recently proposed European "Right to be forgotten" (right to be forgotten: people have the right to request the removal of negative information about themselves or outdated personally identifiable information), etc., will also make public opinion representatives accountable to administrative agencies and to investigate those who sell technology and machines. It is more difficult for private companies. Both of the foregoing have caused profound damage to democratic politics and human rights.

When adopting a technology, one should first realize that the governance of a country or a city is essentially a "political" activity and should not be a purely technical choice. In the process of development and deployment, we should constantly ask (yes, "constantly"): " Who is good for? Who has a voice? Who can supervise? It will give those people or units more power. ? ” Only continuous monitoring and checks and balances can protect and promote democratic values and democratic polities.

5. Cultivate the ability of municipal units to use data and formulate procedures

Apart from the conflict of interest between the government and the people, selfishness and lack of rectification within the executive branch are often the main causes of ineffectiveness (rather than access to the latest technology). The city government or the central government is full of various units, and each unit often has its own independent system, file format, data classification or processing procedures, which makes the analysis work difficult, not to mention seeing through the performance and avoiding "false correlations". The degree of cooperation and willingness to communicate between people cannot be solved by computers. If readers are a little older, you may remember that in the early days, each county and city had its own free WIFI, but each county and city had to apply for a different account password for free WIFI. After several years, it was unified into "iTaiwan", and it was only possible to register once for the whole Taiwan.

Another part that deserves review is the resistance of the personnel themselves to new technologies or techniques. This is the most obvious in Taiwan's tax reporting system: the IRS and the Ministry of Finance already have the most complete financial details of each individual, but they still require people to spend a lot of time reporting it themselves. Even if online reporting is possible, the inhumanity and unfriendly nature of the system interface is always a common problem for government agencies when developing applications or websites.

In addition to the above five points put forward by the author of "Smart City Kidnapped by Technology" , I personally think the following points should be added:

Return to cultivating and improving enthusiasm for participating in public affairs

Taiwan has begun to attach importance to information technology education, and even wants to include AI courses from elementary school (although I really think it is crazy); the purpose is all for the so-called "enhancing industrial competitiveness" or "enhancing employability" and other reasons. However, when it comes to participating in public affairs, most parents or school teachers still hold the mentality of "as long as you study hard and don't care about the rest". In other words, if there is no change in people's attitudes in the field of civic literacy, then even sending a copy of the source code of the algorithm or the blueprint of the supercomputer to students from doctoral classes to kindergarten will still not help the municipal government "hidden by technology" or the liberation of state affairs. People's indifference and irrelevant attitude towards politics and mass affairs, and the bureaucratic system that does not want people to see too much and ask too much, also have a serious lethal effect on democratic politics.

The above is my book review based on the book "Smart City Kidnapped by Technology" published by Pedestrian Culture Lab in August 2020. May good books help you and I grow together.

book information

"Smart City Kidnapped by Technology"