In the AI hype, you might forget that artificial intelligence has been deployed by organisations for years. There are already important lessons to be learned. AI can certainly help make an organisation more efficient, but we are already seeing how AI can have a negative impact on employees.
Using AI, more complex management tasks, such as monitoring employees and assessing job applicants, can now be taken over by machines. Although still in its early stages, algorithmic management - delegating management functions to (AI) algorithms in an organisation - is becoming an important part of digital transformation in organisations.
AI makes processes more effective and efficient
Algorithmic management promises to make work processes more effective and efficient. For example, algorithms can speed up the hiring process of employees by filtering large numbers of applicants at relatively low cost. Algorithms can also enable companies to understand or monitor employee productivity and performance.
Algorithms control the robots in a fully automated warehouse where goods are placed 'haphazardly' at first glance. The algorithms learn where a box or pallet can best be placed and picked up again most quickly. Other algorithms discover much earlier than a manager which employees are considering resigning.
AI does not necessarily make work better
When implementing algorithmic management, however, ethical challenges and potential negative disadvantages for employees must be taken into account. In the case of hiring, AI tools have faced heavy criticism due to harmful biases that may disadvantage different groups of people, which has led to efforts to establish guidelines and regulations for ethical AI design.
Research shows that the exclusive focus on efficiency can reduce employee satisfaction and performance in the long run by treating employees as mere programmable "cogs in a machine". This can significantly undermine employee wellbeing and satisfaction, for example by encouraging employees to keep working until they are exhausted. Think of Amazon warehouse workers and Uber drivers.
Big AI is watching you
AI is very capable of watching people. They know this in China, but organisations in the Western world are also deploying algorithms to monitor the performance of their employees. This is ethically very questionable and it also leads to resistance from employees.
People who know they are being watched adjust their behaviour. This leads to a lot of stress and, in some cases, business accidents. This is not only undesirable, it also damages an organisation's reputation and leads to a lot of staff attrition.
Symbiosis between AI and human work
The researchers behind the MITSloan article Algorithmic Managment: The Role of AI in Managing Workforces argue for more symbiotic relationships between what algortimes do and what managers do. Algorithms are fine for small tasks that repeat endlessly. But human managers are much better at overseeing all the factors involved in managing employees.
That seems like an open door, but in practice, management struggles with the optimal use of AI. Expectations are often too high, the damage of improperly deployed AI is high and the disappointment is then so great that people want to abandon AI altogether. Again, this is not helpful when competitors may be using AI optimally.
AI as a management aid
Look at the below illustration to explain how AI can help management.
At lower levels of management in an organisation, AI systems can provide task coordination while human managers oversee and collaborate. At higher levels of management, AI scans the environment to provide information and decision support, while human managers bring a strategic and holistic perspective to decision-making.
Algorithms are human work, with all the mistakes that go with it
At the end of the article, the authors warn about the dangers of biases in algorithms. We already know the cases where women or people with non-white skin colour were discriminated against by the algorithms.
What they do not mention are the challenges for privacy protection and problems in protecting sensitive business information. Those risks should not be underestimated.
But AI and algorithms are here to stay. We clearly need to learn how best to use this technology. The benefits are many, but it is still important to be vigilant.
The MITSloan article Algorithmic Managment: The Role of AI in Managing Workforces provides further examples of the deployment of AI in organisations.
