Management by Algorithms

Algorithmic management refers to the use of algorithms to monitor and control the activities of human employees. As algorithmic management becomes increasingly widespread, it is critical to understand what it is, how it works, what are the benefits and drawbacks of using management by algorithm, and what the future holds for this approach. We will explore what are some of the innovative companies employing algorithmic management in their core business operations and their perceived benefits and drawbacks

Introduction

Since the advent of a "gig economy" in recent times, how way thousands around the world operate has changed dramatically. Some argue that it increases people' flexibility and opportunity while also removing obstacles to employment, while others argue that it weakens organizational guidelines and values while encouraging businesses to see workers as expendable.

The gig economy will undoubtedly continue to exist and thrive. However, with an increasing number of workers working in flexible and freelance roles, how can companies guarantee that they are handled effectively?

Taking use of advancements in artificial intelligence, "algorithmic management" refers to the practice of using programs to monitor the actions of human workers. Since algorithmic leadership is becoming more prevalent, it is important to grasp what it is, how it works, the advantages and disadvantages of its use, and the future of this field.

The algorithmic management mechanism

Algorithmic management is all about using automated systems and evolutionary algorithms to administer a collection of social employees. The goal is to automate significant portions of the management decision-making process via the collection and analysis of massive quantities of data, most notably data on employee performance.

There is evidence to suggest that algorithmic management is gaining traction. Over 50% of human resources departments of global companies are said to be using artificial intelligence technology at the moment. (Dominican Republic, 2020) Here are a few instances of how algorithmic management is making inroads into the mainstream:

1. HiredScore is in the process of acquiring artificial intelligence software that evaluates a candidate's expressions, manner of speech, and linguistic use during video interviews. (Nothing, 2021) They assert that the new approach may significantly shorten the recruitment process, while others assert that it will aggravate already-existing socioeconomic inequalities and injustices.
2. RedMart's Singapore warehouse is managed by algorithms that determine which items need to be gathered, moved, and stored prior to being shipped out. The "pick rate," a metric that indicates how many items are taken from the shelves each hour, is believed to put pressure on workers to improve their performance in this area. (2021, Pearly)
3. Grab Drivers for such grocery chain obtain financial individualized performance reports that detail their performance, including the mean time it would take them to fulfill calls, the mean time it takes them to travel to diners, the mean time it takes them to travel to consumers, as well as the quantity of delayed or unallocated orders. Drivers must accept new customer orders in an average of 30 seconds, which is much quicker than the industry norm. (Singapore Straits Times, 2019)

The benefits and drawbacks of algorithmic management?

Algorithmic management and stochastic commerce provide up new opportunities and efficiency for companies and their employees. Several potential benefits of algorithmic management include the following:

1) Decisions based on information: When you use algorithmic management, you may make more objective decisions than if you relied on unscientific "gut feelings."

2) Reduced labor expenses: By outsourcing at least a portion of your managerial duties to an algorithm, you may significantly decrease your organization's labor expenditures. Operations that would take human hours to perform may be accomplished in seconds by a powerful computer using artificial intelligence.

3) Improved Efficiency: The application of algorithmic management may assist managers in more efficiently scheduling employee shifts and assigning assignments, resulting in increased worker productivity and reduced wasted time for both managers and employees.

4) Increased Objectivity: Algorithmic management may also help mitigate or eliminate human bias and partiality in decision-making. The underlying assumption is that the algorithm was designed without biases.

Despite these benefits, algorithmic management is often applied in ways that are controversial, if not downright harmful, to an organization's objectives.

According to the results of a 2019 study of Grab drivers, the majority of their concerns fell into three broad categories, each of which emphasized the issues connected with algorithmic management. (Singapore Straits Times, 2019)

 

1) Surveillance and observation

Drivers are aware that they are constantly watched by the Grab smartphone application, which tracks important performance indicators including vehicle velocity, Location data, and new passenger acceptance rate. (Singapore Straits Times, 2019) It is possible to be punished or even permanently banned from the site if you participate in activities that do not align with the KPIs.

Grab's aim of collecting as much data as possible makes sense, considering their inability to place an actual director in the backseat for each ride they offer. However, efforts to monitor employees in order to boost productivity may sometimes backfire, leading in decreased engagement, morale, and a breakdown of trust among employees. Several major companies, including Barclays, recently abandoned plans to develop monitoring tools that might track how much time employees spent at their workstations and what proportion of their time was spent on each task. (2021, Mark)

2) Artificial Intelligence in a Blackbox

Several Grab drivers have voiced worry about a perceived power imbalance between the smartphone application, which constantly monitors their productivity, as well as the workers, who have limited knowledge of where the business operates.

Grab asserts that revealing excessive details about the algorithm will endanger the company's revenues since the program's internal dynamics constitute a trade secret and may be compromised. The algorithm may be so complex and dynamic, always adjusting to changing circumstances, that even technical experts may struggle to explain what is happening in the background of the software. (Singapore Straits Times, 2019)

Regardless of how valid these reasons are, the fact is that computational governance may be excruciatingly opaque to the workers who are subject to its demands and instructions. Due to the fact that the human brain despises ambiguity, employees who are not provided with sufficient information may feel disrespected or mistreated. So according to yet another psychological research, when employees feel "out of the loop," their evaluation of their group's standing falls by 70%. (2017) (Eric, 2018)

3) Silos of Humanity

Perhaps more upsetting than the lack of openness itself is the resulting feeling of isolation. The majority of gig workers lack coworkers or supervisors with whom to interact; instead, they have just themselves, the smartphone, and customers who come and go each day.

Without establishing a relationship with a human boss, drivers may struggle to understand their own performance or feel they are carrying out a significant job. While the issue is particularly severe for Grab drivers, the sense of dehumanization may afflict anybody who is exposed to algorithmic management in any position. Individuals risk feeling disconnected and alienated from their coworkers as technology becomes more ubiquitous in the workplace.

When adopting "just-in-time" scheduling, automated shift scheduling software is utilized to guarantee that employees arrive and depart from work at the precise times necessary to accomplish given duties. On the other side, practices such as just-in-time scheduling have been shown to increase strain, income volatility, as well as the probability of employment conflicts by disrupting workers' schedules.

Conclusion

Due to the potential disadvantages discussed before, businesses considering in implementing algorithmic management may adopt measures to improve their chances of success.

One of the reasons algorithmic management is so disliked is because communication between the algorithm and the worker is one-way. By including workers within judgment process, it is possible to boost employee engagement and, in certain instances, even improve the algorithm itself.

Businesses that use algorithmic management should make an effort to show their care for their workers' well-being, rather than considering people as COGS in a machine. Certain businesses in the United States have pushed for a variety of labor reforms, such as a strong federal sick leave fund for autonomous employees and a countrywide transferable standard that ties compensation to the employee rather than the company.

Even in circumstances when workers are directly overseen by an AI program, human interaction may make a big impact. Organizations should retain human resource management and support roles and offer tools to allow employees to develop connections with their colleagues. AI agents that can adapt their behavior, make autonomous decisions, and evolve independently of human involvement. The ability of AIs to generate additional Artificial intelligence, in particular, would need a rethinking of organization structure, strategies, and leadership models that don't really help distinguish between AI and human agents. When it comes to algorithmic management, organizations will be required to maintain a high level of transparency.

Author: Tan Kim Chong

 

Bibliography

Dom Nicastro, AI Is Reinventing HR (May 2020)

Nada Chaker, The Missing Link in Talent Transformation (March 2021)

Pearly Neo, Unbroken Cold Chain (May 2019)

Straitstimes, Regional Super App, (Sept 2019)

Mark Murphy, Barclays Stops Employee Tracking After Backlash, (Feb 2021)

Eric Jones, Who Is Less Likely to Ostracize? (June 2018)

Andy Williams, Artificial General Intelligence Model (July 2020)