Who can help me dance with algorithms—Leadership in algorithmic management

With the rapid development of digital technology, the workplace is undergoing earth-shaking changes; algorithms are increasingly favored by enterprises and have become an important role in promoting management innovation. It is worth noting that when some companies actively advocate "algorithm management", they relatively ignore the value of people and encounter many new troubles. Algorithmic technology was originally invented to improve human welfare, but unfortunately it has become an upgraded tool to "enslave" people. This reminds leaders that they must correctly examine the “pros and cons” of algorithmic management.

The "Pros and Cons" of Algorithm Management

Lindebaum et al. (2020) proposed that an algorithm is a series of well-defined finite methods used to solve specific problems in batches. Step or sequence instructions. Algorithms enable companies to complete complex tasks that were previously impossible with human or simple electronic technology alone. For example, the online car-hailing platform manages the limited resources of the online car-hailing system based on scheduling algorithms to ensure rapid and reasonable resource allocation to best meet the needs of drivers and passengers. Manufacturing companies equip heavy machinery operators with fatigue detection headbands and badges with GPS positioning functions to remind employees to conduct safe and compliant production operations to prevent accidents, and incorporate these safety-related performance data into the final assessment . In addition, some companies also use algorithms to carry out gamification task design to provide real-time feedback and rewards for employees' appropriate behaviors, improve employees' work enthusiasm, and encourage them to be more willing to invest in subsequent work. The above examples fully reflect the positive role of algorithms in improving management efficiency. With the support of computing power, the algorithm utilizes or generates a large amount of data, including employee task performance (behavior), physiology, psychology and real-time location. Organizations or leaders can use these multi-modal, multi-category, and large-scale data to assist management decisions, thereby improving management efficiency, helping employees complete their work more efficiently and easily, and improving employees' sense of work meaning and personal happiness.

But on the other hand, algorithmic management is also regarded as "Taylorism 2.0". Similar to Taylorism, the core of algorithmic management is algorithmic control, which improves the overall efficiency and performance output of the organization by strengthening on-site data collection, process analysis, and efficiency management. Because algorithms can help companies achieve broader, finer-grained, more real-time, and more influential control over their employees, many companies will inevitably be coerced by instrumental rationality and want to continuously highlight the advantages of algorithms and replace people in management decisions. Neglect to consider the complexity and applicability of digitalization itself, as well as the changes that occur to organizations, managers, and employees after the introduction of digital technology, as well as the coordination and adaptation issues among them. As you can imagine, the results of doing so are often counterproductive. Algorithmic management drives companies into a Taylorist dilemma, and ultimately does more harm than good. Take the food delivery industry, which is common in life, as an example. Driven by algorithms, delivery workers are involved in an invisible competition. In this competition that will never end as long as there are people participating, the algorithm plays the role of a rational, cold, and powerful controller behind the scenes, and it also mobilizes customers to impose broader constraints and incentives on delivery workers through comments. Therefore, takeaway workers are subject to system rules and logic and have to race against time to increase order quantity and delivery speed, and even violate traffic rules in order to deliver food on time, threatening their lives. It is conceivable that under the constraints of the algorithm, delivery workers are exhausted and their professional identity and personal happiness are seriously damaged. This kind of plot is not only played out in the food delivery industry, but also in various traditional industries where a large number of employees are trapped in the algorithm.

Individuals under algorithmic management: Human nature is evil or human nature is good

In response to the key question of "how to better leverage the technical advantages of algorithmic management and reduce its negative impact", leaders One must first ask “who are the people directly affected by algorithmic management?”

Mc Gregor (Mc Gregor) proposed Theory X and Theory Y about human nature assumptions in "The Human Aspect of Business". Specifically, Theory In this case, companies need to use a "carrot and stick" approach, that is, to supervise and motivate employees through strong controls. Algorithmic management can replace traditional regulators in terms of control. For example, algorithms can use technologies such as pattern recognition and natural language processing to automatically analyze employee behaviors in video clips and further infer employee psychology. This level of automation and intrusion is unattainable with traditional regulatory methods. Following the logic of Theory All in all, if employee motivations are as presupposed by Theory X, organizations should treat employees as depersonalized machines and maximize the role of algorithmic management in exercising tight control over employees. It is worth noting that some companies use algorithms to exercise excessive control over employees. For example, they use algorithm systems to obtain employees’ online browsing records and behaviors to predict employees’ turnover intentions and likelihood of resignation in advance. As another example, other companies use "latent nets" similar to what Breed described in "Social Control of Editorial Offices" (using communication media in a specific social environment to achieve invisible but powerful social control) to Employees conduct silent surveillance. With the support of algorithms, everything from employees' speech and behavior on the company's intranet to employees' social media usage records outside the workplace may become tools used by hidden networks to exert control over employees.

Theory Y, which is opposite to Theory Get up, exercise self-discipline and demonstrate appropriate behavior in the workplace. External control and punishment can threaten and hinder employees. If we follow the logic of Theory Y, imposing strong algorithmic monitoring may cause employees to feel that their work autonomy is restricted and their privacy is violated, which in turn will reduce employees' trust in management and the organization, and ultimately harm employees' work performance and personal happiness. feel.

In fact, Theory X and Theory Y do not portray an absolute distinction between good and evil, but reflect the dual functions that algorithmic management may play on individual employees. It is not difficult to find that Theory Y has a relatively more benign understanding of human nature and employee work motivations. The author believes that although companies need to exert appropriate control over employees in order to maintain efficient operations, companies should never neglect to pay attention to employees as unique individuals in the management process. In the long run, a company's greatest asset is its talent. Only by acquiring, retaining, developing, and making good use of talents can enterprises ensure sustainable development in a fiercely competitive environment, especially in the wave of digitalization. Through algorithmic management, managers can obtain real-time multi-modal big data collection capabilities that cannot be achieved through traditional methods; this ability can help managers better understand employees and achieve efficient decision-making with employees. People must realize that even if algorithmic management has caused a certain degree of "de-individuation", "achieving individuals" is still one of the ultimate goals of organizational development.

We urgently need algorithmic leadership

Algorithmic management is a double-edged sword. If the introduction of algorithmic management is regarded as the will of the top management of the organization, then companies should incorporate more human considerations when designing algorithms. For example, in order to maximize labor value and improve efficiency, food delivery platforms continue to shorten the delivery time of food delivery riders, resulting in the emergence of many labor conflicts. In response to the crisis, food delivery leaders Ele.me and Meituan have both tried to optimize their algorithm systems to give riders more flexible time and proactively correct social deviations in technology to a certain extent. At the same time, people must realize that technology and people are not isolated. Technology should be used as an auxiliary and strengthening tool to help managers make decisions, rather than replacing the leadership role of managers in organizations and teams. Therefore, when enterprise managers design or introduce specific algorithms to promote digital transformation and improve efficiency, they must also think about how to use their own initiative to deal with potential risks caused by algorithm management. This involves the emerging issue of algorithmic leadership, that is, in the context of algorithms, what emerging challenges will leaders have to deal with, and how to effectively help employees "dance harmoniously" with algorithms.

Specifically, leaders must be clearly aware that employees face many "tensions" that cannot be ignored under algorithmic management, between employees and machines, organizations and colleagues. First of all, while artificial intelligence and automation liberate employees from mechanical repetitive work, they also cause employees in some traditional positions to become technologically unemployed or suffer greater pressure from skill requirements. Faced with these substitutions and threats, employees will naturally feel uneasy: Will I be replaced by a machine? What is the difference between me and the machine? Why should a machine manage me? As a leader who possesses authentic intelligence rather than artificial intelligence and has a human touch, he needs to respond to the subjective tension between employees and machines.

Secondly, organizations use algorithms more to reduce costs, increase efficiency, improve quality, and innovate, often putting employee development and interests in a secondary position. In this process, there may be situations where the organization sacrifices the individual interests of employees for the overall interests. For example, factories issue "optimal" action instructions to workers based on efficiency maximization and algorithmic calculation results. However, this help may conflict with workers' own experiences and habits, causing them to experience role ambiguity and reducing work autonomy. As an important link between the organization and employees, leaders need to respond to the interest tensions between the organization and employees.

Third, algorithms may intensify competition and distrust between people. Under algorithmic management, employees are materialized into machines that produce data, which further affects employees’ views of themselves and their recognition of their colleagues around them. In this case, do employees view their colleagues as partners working side by side, or as rivals competing against each other under the baton of the algorithm? As significant others who exert direct social influence on employees in their organizations, leaders need to respond to collaborative tensions among employees.

De Cremer, a well-known scholar in the field of human-computer interaction behavior research, proposed in the book "Algorithmic Colleagues" that compared with traditional managers, leaders in algorithmic management need to accept the continued Education and training, using technology-led efficiency models as new ways of working, and helping employees understand and utilize new technologies; at the same time, leaders need to pay attention to their own uniqueness as human beings (such as creativity, empathy, ethical judgment, etc.) and adopt Establish a good social relationship with employees in a more proactive and emotional way. It can be seen that leaders in algorithmic management must know how to find a balance between technology and human nature.

In order to give full play to the role of leadership in eliminating the adverse effects of algorithms on employees and promote efficient collaboration and common development between people and algorithms in the organization, the author believes that algorithmic leadership should have three typical behavioral connotations, namely Humanistic care, goal trade-off and relationship synergy.

#First of all, algorithmic leadership should be a leadership that emphasizes humanistic care. Some argue that among the core functions of leadership, only deep interaction, proactive representation and leading change are not in danger of being immediately replaced by machines. Algorithmic management has taken over some of the functions of real leaders to a certain extent, reducing the opportunities for face-to-face two-way communication and understanding between leaders and employees, and reducing the depth of interaction between leaders and employees. The work connection and social connection between leaders and employees have an appropriate role in establishing a good superior-subordinate relationship and improving employees' role performance and situational performance. Therefore, the role of algorithmic management does not mean that leaders are absent from management. Leaders should play the role of humanistic care and eliminate the adverse impact of algorithms as a cold supervision and evaluation subject on employees.

Leaders with humanistic care need to deal with the problem of helping employees meet their "emotional needs" while accepting the "rational management" of algorithms. Leaders need to let employees know that algorithm management is negotiable, that organizations and teams value employees' reasonable demands and work experience, and that the purpose of applying algorithms is always to help the organization and employees better improve efficiency, rather than to control work. At the same time, leaders also need to pay attention to the uniqueness of each employee under algorithm management and timely integrate into the interaction between the organization-algorithm-leader-employee, instead of just treating employees as machines that continuously produce data, in order to dehumanize and dehumanize them. Treat employees as equals.

Secondly, algorithmic leadership should be a leadership that focuses on goal trade-offs. Algorithms are designed and introduced at the organizational level to use digital technology to empower corporate management processes and improve corporate efficiency and market competitiveness. In contrast, the basic demands of employees are the satisfaction of personal interests, such as obtaining considerable compensation, promotion opportunities, and self-realization. In the context of algorithmic management, it is often difficult for companies to balance the interests between the organization as a whole and individual employees, and algorithms, as a means of controlling the employee labor process, enable companies to gain a greater advantage over employees, and often Choose to sacrifice the interests of vulnerable employees when there is a conflict of interest. Therefore, leaders at the intermediate level between the organization and the grassroots employees have the responsibility to think more about how to play their role and effectively respond to the tension within the organization, especially between the organization and employees, such as "profit maximization" and "profit maximization" The tension between "humanizing management", "organizational quantitative calculation" and "employee value proposition". This first requires leaders to pay attention to the human nature of the employees they manage, have warmth, be good at detecting various emotional expressions of employees, listen to and give timely feedback to employees' demands under algorithm management. In addition, leaders must give full play to the bond between the organization and employees, be good at balancing individual interests and collective interests, and do their best to find a balance between the goals of different stakeholders.​

Finally, algorithmic leadership should be a leadership that advocates relationship collaboration. To a large extent, algorithm management realizes the guidance, evaluation and discipline of employees' individual labor process, and can provide real-time and all-round monitoring and performance feedback to employees. This may lead to employees paying more attention to their own performance goals and labor processes and results, to a certain extent, neglecting how to cooperate with others in the team and organization to achieve set goals. Algorithm management relies too much on structured digital workflows, resulting in a significant reduction in human interaction, resulting in low-quality collaborative relationships, and even becoming a "disorganized mess", which is not conducive to effective cooperation. In today's society, the proportion of knowledge-based employees is gradually increasing. They have a strong desire to realize self-worth, attach great importance to achievement incentives, and are keen to fully demonstrate their unique competitiveness through refined evaluation results. In some jobs where processes and results are relatively easy to quantify, algorithms may further deepen the tendency of knowledge workers to work alone. However, for the teams where knowledge workers work, the lack of open collaboration and knowledge sharing will ultimately affect the performance of the entire team. progress.

It is true that people should consider indicators of cooperation process and performance at the beginning of algorithm design, but leaders still need to play their own role to effectively reduce the alienation and self-interest among employees caused by individualized evaluation. tendency. Leaders can make full use of the online communication channels provided by digital technology to promote efficient multilateral communication between themselves and their employees, keep abreast of employees' work, life, and family situations, and give sufficient feedback to help employees better integrate into the team. , experience the support of the team, leaders and colleagues. Leaders can also actively use methods such as visualization of team and individual work progress, outdoor team building, knowledge sharing meetings, etc., to allow team members to better internalize the overall team goals into their own personal goals, and on the other hand to strengthen the relationship between team members. Understand each other, reduce the occurrence of distrust and misunderstandings, and at the same time strengthen the transmission of information between each other, so that employees in the team know the relative strengths and weaknesses of each person, thereby increasing the willingness to cooperate with each other in the process of completing the work, and better Seek to complement each other’s strengths and improve work efficiency and quality.

Conclusion

In short, algorithm management is a double-edged sword. As tools, algorithms help companies achieve previously unimaginable scope, depth, and intensity of management. But the tool itself should not replace the purpose. The purpose of algorithmic management is always to improve efficiency rather than control, and algorithmic management should not replace human management.

In the context of algorithmic management, the value of people - whether it is an employee as a person with emotions and motivations, or a leader as a person who collaborates and complements the algorithm - is always the specific aspect of algorithmic management. What should be considered during the application process. Algorithmic leadership, whether it is a new concept or a new connotation derived from traditional leadership in the context of algorithms, is worthy of active connotation exploration and practical application by corporate managers in the context of real algorithm applications.

The above is the detailed content of Who can help me dance with algorithms—Leadership in algorithmic management. For more information, please follow other related articles on the PHP Chinese website!