[Worker** Editorial].
Original title: To guide algorithms for good, more effective supervision is needed.
Worker**-Wu Di, commentator of China Labor Network.
In the age of the Internet, we can live in algorithms, but we can't live in calculations. As an algorithm is a technology that serves people and society, algorithm governance should ultimately be implemented to protect the rights of users. We look forward to the active participation of the national top committee members and industry insiders to put forward more valuable suggestions and exploration directions.
Online car-hailing users who often accept a long distance will have a high probability that the car shown in the next dispatch order is quite far; If you can't call an online car-hailing car, you often add money, and the chance of calling a cost-effective car seems to be reduced; Users who brush short ** seem to be able to endlessly brush their favorite content ......According to a recent report by The Paper, the 2024 National People's Congress and the National People's Political Consultative Conference kicked off, and some members of the National Committee of the Chinese People's Political Consultative Conference put forward a proposal on improving the governance system of artificial intelligence algorithms on the basis of research, emphasizing the importance of improving the supervision of algorithms.
The more expensive the online car-hailing is, the more you swipe, the more you can't stop ......The public is no stranger to the algorithm behind it. Previously, "big data killing" was many people's personal feelings about algorithms. For example, the goods or services seen by old customers on the platform are more expensive than those of new customers, and searching for services on the same platform with two mobile phones at the same time shows different prices. At the same time, another typical slot of the public about the algorithm is "the takeaway brother is trapped in the system", which once appeared on the social **hot search list.
In the era of artificial intelligence, many people have to accept algorithms in various application scenarios. In the midst of some doubts, complaints, and discussions, the relevant parties have introduced measures, such as clarifying that the principle of users' informed consent is required for information collection and processing, and restricting excessive claims by platforms, and some platforms have also introduced options for users to turn off algorithmic services, etc. These are all aimed at restricting the platform's permission to use data, protecting user information security, and guiding algorithms to "be born for good". However, on the whole, public dissatisfaction still exists, especially with the global artificial intelligence boom triggered by ChatGPT and SORA, and the algorithm technology has entered the application stage in an all-round way, such as algorithm discrimination, algorithm black box, information cocoon, algorithm hegemony and other issues need to be paid more attention to by us.
First of all, the algorithm itself is quite complex, and it is not easy to sort out the mechanism and create relatively fair rules. As mentioned in the above news, "there are hundreds of millions of parameters in the backend of online ride-hailing, and it is difficult for even technicians to adjust which parameters can achieve a fairer result" - similar problems are widespread on various apps.
Second, for new things, policies and regulatory provisions have a certain lag, and most of the relevant laws are frameworkd and advocacy principles, which often leads to insufficient correspondence between the content of algorithm governance and illegal infringement.
Thirdly, in actual cases, it is difficult to identify and protect rights of algorithmic infringement. Algorithm providers have information resources and technological advantages, and it is "difficult for regulatory authorities to understand their specific mechanisms and decision-making processes, and it is difficult to obtain evidence", and then it is "difficult to identify" law enforcement.
In addition, the rapid development of generative AI technology will inevitably bring new challenges, such as some technologies can learn and iterate autonomously, while the existing regulatory system and regulatory technology are difficult to cover in a timely manner, resulting in a regulatory vacuum.
It is precisely because it is difficult for algorithm technology to be understood and effectively managed, and its operation password is in the hands of professional Internet companies, so how to improve regulatory measures, improve regulatory efficiency, and stimulate algorithms for good is a major challenge for regulators.
To this end, in recent years, relevant parties have continued to explore governance measures. For example, in 2021, seven departments including the State Administration for Market Regulation jointly issued the "Guiding Opinions on Implementing the Responsibilities of Online Catering Platforms and Effectively Safeguarding the Rights and Interests of Takeaway Delivery Workers", which clearly stated that "the algorithm is selected", requiring reasonable determination of assessment factors such as order quantity and ** rate, and appropriate relaxation of delivery time limits; "Promote platform enterprises to legally standardize employment, scientifically adjust algorithms, and improve labor quota standards" was written into the report of the 18th National Congress of Chinese Trade Unions; There are pilots where food delivery platforms have launched relatively loose restrictions on food delivery riders, and so on.
However, in order to reduce the negative impact of algorithm technology at the root, it is necessary to further improve the governance pattern. This is also the consensus reached by the national top watch committee members and industry professionals in recent years. For example, some deputies and committee members suggested improving the current joint conference system for algorithm governance with the participation of departments such as the internet information and industry and information technology; Participate in the regulatory process with the help of market expertise and make up for the shortcomings of regulatory technology; explore the use of artificial intelligence technology to strengthen the "technical defense" and "technical supervision" of algorithms, and improve the efficiency of supervision; Let the platform enterprises explain the logical process of the algorithm clearly, and balance the factors such as technology and ethics, and social impact; Expand the path of rights protection, innovate the mode of responsibility determination, and enhance the ...... of users' rights protection
In the age of the Internet, we can live in algorithms, but we can't live in calculations. As an algorithm is a technology that serves people and society, algorithm governance should ultimately be implemented to protect the rights of users. We look forward to the active participation of the national top committee members and industry insiders to put forward more valuable suggestions and exploration directions. It is our common pursuit to guide algorithms to "live for good" and amplify the positive value of algorithm technology with more professional and effective governance measures.
*: China Industry Network - Workers**).