Why regulate the "algorithm"?

The so-called & ldquo; Algorithm & rdquo; It is the essence of computer processing information. In mathematics and computer science, algorithm refers to a defined finite step or sequence of instructions that can be implemented by computer. It is often used in calculation, data processing and automatic reasoning. In short, through the algorithm, you can get a result from an input data. Of course, this result is not necessarily certain. In Internet services, algorithms usually represent some content provided by Internet service providers to users, such as personalized push, sorting and selection, retrieval and filtering, scheduling and decision-making, etc< br />
So why regulate the algorithm? The reason is very simple. First, in the general direction of society, the algorithm directly determines what information internet users can get. As long as people who have used the Internet know, the Internet has a huge amount of information stock and is generating a huge amount of new information all the time. For Internet users, it is impossible and unnecessary to know all the information. Users are usually only interested in their preferred information, so the information obtained by users must be screened. In the paper media era, this screening was carried out manually by newspaper editors. In the early days of the Internet, this screening was also carried out manually by editors of portal websites. But now, especially in the era of we media, it has long been difficult for people to deal with such a huge amount of information on the Internet. Therefore, a large number of Internet content providers often decide what content to provide to users through certain algorithms. In this sense, what kind of information users can obtain is completely determined by the algorithm. Moreover, because the algorithm is established on a certain platform, the information screened by non manual automatic algorithm is easy to cause & ldquo; Information cocoon room & rdquo; The effect of. On the one hand, the audience will only receive similar information, and it is easy to fall into their own & ldquo; Stratosphere & rdquo; I can't extricate myself from it. If only one kind of information is accepted for a long time, it will not only seriously affect the correct formation of the audience's outlook on life and world outlook, but also constantly amplify the views echoing in the stratosphere and go to extremes. On the other hand, this & ldquo; Information cocoon room & rdquo; The formation of the algorithm will also leave space for others to manipulate information. As long as some specific content is added to the algorithm, there will be a specific tendency in the result of the algorithm, and the audience will be inadvertently affected by this specific tendency, resulting in some specific views. This kind of operation is common on some Internet platforms at home and abroad. Under specific conditions, it will have a significant impact on the whole society< br />
On the personal side, the regulation of algorithms is also closely related to the protection of personal privacy. No matter how clever the algorithm is, it is only a program after all. It can't get results without inputting content. In personalized push, sorting and selection, retrieval and filtering, scheduling and decision-making and other Internet services, the main content that the algorithm needs to collect is the user's browsing habits. Generally speaking, today's algorithms can judge users' personal preferences through users' browsing records, what content users have browsed, and how long they stay on various contents in the browsing process, and provide users with relevant content according to users' preferences, so as to tempt users to stay on relevant platforms for a longer time. The user's browsing habits are closely related to the user's private life, which undoubtedly belongs to the user's personal privacy. Now, in order to ensure that they accurately push content to users and increase user retention, many Internet service providers will set the application of consent algorithm as the premise of receiving services. If users want to install or use relevant software, they must allow the software to collect relevant browsing information. Therefore, these Internet service providers have a lot of user privacy, which greatly increases the risk of user privacy disclosure< br />
Because the algorithm has such risks, from the perspective of maintaining social order and protecting personal privacy, the algorithm should be properly regulated, so that the algorithm can not only provide for users to use Internet services, but also try to limit the negative effects of the period. In this regard, the Internet Information Office recently published the regulations on the recommendation and management of Internet information service algorithms (Draft for comments), which mainly requires Internet service providers to & ldquo; Inform & rdquo; in a prominent way; The existence of the algorithm and & ldquo; Publicize in an appropriate way & rdquo; Basic principle and operation mechanism of the algorithm (Article 14). On this basis, the exposure draft requires service providers to provide users with non personal targeted choices, or provide convenient options for modifying and closing algorithms. That is, in terms of algorithm regulation, the exposure draft clearly stipulates the notification principle and the user's exit mechanism (Article 15). At present, this approach is the most possible way to balance the interests of all parties< br />
Although various risks of the algorithm are mentioned above, if the algorithm is completely excluded, it will not only affect the income and profit of Internet service providers, but also affect the user experience, which is not conducive to the further development of the whole Internet industry. On the premise that Internet service providers fulfill their reasonable obligation of disclosure, the user has the right to choose whether to turn on the algorithm, that is, let the user judge whether he is willing to transfer part of his privacyIn exchange for a better user experience and more useful information, users will naturally choose software and platforms with less risk and better service experience. In order to retain users, market competition will promote relevant enterprises to develop better algorithms and provide users with better service experience while minimizing privacy collection, which is more conducive to promoting the benign development of the whole industry< br />
It should be said that there is no precedent in the world to try to regulate the algorithm. Naturally, this practice in China can not be achieved in one step, which requires everyone to pay attention to this problem, because the algorithm is indeed closely related to each of our daily life. Enterprises and the whole society participate in this legislative process and seek a solution acceptable to all parties in the spark of collision, the most realistic and possible normative framework.