I attended the AI Paris exhibition on June 11 and 12 and was able to attend several interesting conferences. At 10am David Gibal and Jean-Christophe Gard start talking to us about AI and the future of work, by presenting an exclusive study: the Malakoff Mederic Humanis study. This speech particularly caught my attention, considering that several very interesting points were raised, including the importance of HRD in the face of artificial intelligence.  

Malakoff Mederic Humanis study

This barometer is called “AI and human capital: humans are the future of AI”. The main issue being: “what are the impacts of AI on the organization of work, the nature of jobs and human capital?”

We can already notice some paradoxes: 69% of managers consider that AI will strongly impact the company, yet only 18% place it as a strategic subject for their company. Therefore, developing AI is important, but it is not a near strategic issue for executives.

This barometer reveals that not all levels of the company have the same perception of AI: for some it is a threat, for others it is an overpowering machine for example. It is therefore important to educate this population on the subject, in order to obtain an optimal use of artificial intelligence in the company.

However, this role, which can be likened to the management of change in the company following the installation of AI, falls to the human resources department. This mission is absolutely essential, it represents the longest and most decisive step in the installation in artificial intelligence.

It is therefore up to human resources to manage all the human factors of a successful “artificial intelligence project”. According to this barometer, 70% of this success depends on managerial support.

Following findings on the ethical concerns of company members and the different perceptions they may have of the impact of AI, particularly depending on their hierarchical position, the Malakoff Mederic Humanis study cites the 5 missions of human resources that seem essential.

The 5 essential missions of human resources

Skills management

“87% of executives believe it is the role of Human Resources to identify the changes in skills needed and 80% to support company managers in taking AI into account.”

Indeed, the skills needed in a company  will have to change as a result of the entry of AI. The HRD must anticipate this change. This management of skills will mainly involve recruitment or training.

If the use of artificial intelligence is optimal, employees’ tasks will change. The organization of work must therefore be rethought. The tasks will be divided between the employees and the artificial intelligence.

Man and machine work best when they work together. Only intelligent collaboration between HR managers and artificial intelligence will generate optimal value.

Governance and talent capture

The Malakoff study distinguishes three types of activities following the installation of AI:

• Removed activities: these are all activities that can be automated, without added value, and that are repetitive.
• New activities: In this category there are both creative activities, such as algorithm development, data management; but also intermediation activities, such as supervision activities).
• Hybrid activities: hybridization here refers to “the transformation of activities that rely on AI, especially for decision making.” These will be the activities where the AI will act as a simple assistant. For example, AI that helps a cardiologist in his diagnosis, or everything that concerns predictive maintenance…

Faced with this distribution, the HRD will then have to play a role in structuring artificial intelligence. It will be up to her to find and recruit the best talent in the new activities. For example, it will be up to the company to create a department dedicated to data management.

She will also have to reassure the teams about this distribution. According to the study, the majority of employees believe that artificial intelligence will mostly impact the activities that are being eliminated. There is therefore a work of demystification that must be done. Artificial intelligence is an assistant, it will never make the final decision, or act in complete autonomy.

Health and quality of life at work

“Leaders also expect HR to analyze and anticipate the impact of AI on health and quality of work life (83%).”

This mission includes the demystification of artificial intelligence. Indeed, AI can cause an upheaval in the work of employees, both in its content and in its organization. Employees, as a result of automation, may also face cognitive overload if they only manage complex activities at a high pace. It is therefore necessary to accompany each employee in this change, but also to take into account human limits.

However, in theory, automation mainly has a beneficial effect on the quality of life at work: reduction of dangerous tasks, increase in skills, increase in social interaction, improvement in the interest of the tasks…

In the same way, AI can help human resources to train employees, help them with their career path, collect feedbacks… AI allows a better personal follow-up of each employee.

Ethics

“78% of executives today believe it is the role of HR departments to combat ethical biases that could be introduced by AI.”

Human resources must fight against discriminatory biases that can be induced by AI (e.g., algorithmic gender bias). In order to do this, its development must be structured and controlled.

There are several steps that can be taken towards a more ethical AI:

• Regarding the data: A baseline dataset to detect bias can be set up. It is very important to control the data with which you feed the AI.
• Artificial intelligence should not have final decision-making power. It is there to assist the human exclusively.
• The teams dealing with this artificial intelligence must include an explainability part of the algorithms, so that every decision made can be fully justified by real data.

This ethic is essentially based on trust. Thus, this climate of trust must be established early, rehearsed and disseminated. People with access to AI data must be trusted individuals. Once this trust has been introduced, it is up to human resources to make it last, through focus groups, structuring, training, etc.

The creation and sharing of value.

78% of executives consider that AI should be integrated into the social dialogue within the company.

Again, it’s about demystifying AI, but also making it more collaborative. The things we know are less frightening. Only 40% of employees say they are willing to implement decisions made by AI. By better understanding how it works, these employees will have more confidence in it.

In this process of social dialogue, the IRP (institutions representing the personnel) will have an important place. Indeed, they will be able to create an effective link between employees and managers on this subject.

Conclusion

Artificial intelligence, especially when applied to service automation, is a real issue for companies. It provides a better customer or employee experience and optimizes costs.

Most CIOs have taken up these issues and have projects under development (or already deployed). In the early days of this movement, the questions were primarily technical:

1. what are the technological solutions available on the market, what are their differences,
2. what are the most interesting use cases for my organization, which business units will benefit from them
3. how do I proceed technically, do I have enough data, etc…

These projects are already starting to have a human impact, if only on the internal resources needed to complete the project. The logic specific to each of the machine learning domains requires a mental exercise that is not always obvious.

HRDs will therefore have to deal with two issues:

1. how to build an organization where part of the tasks can now be managed by robots,
2. how to develop my employees towards advanced digital skills that will allow them to successfully carry out AI projects

You can also read this article in PDF format here.