Photofit of the meta-employee

Tomorrow’s employees will be assisted, supported and augmented by artificial intelligence (AI). Though AI won’t replace them, it will transform them. A glimpse of what you’ll be, or already are to some extent.

By Kevin Erkeletyan

A CHECKER

“There will always be human verification” This statement by Jean-Luc Leblond (Malakoff Humanis) evokes the twofold role of the meta-employee as an employee that needs to be vigilant and also has an important task: checking. Vigilant, because AI “will go on hallucinating,” says Eneric Lopez, and because it remains an unknown: “we don’t know what’s really going on under the bonnet,” says Luca Paltrinieri. Checking, because one of their new skills will be to check the work done by AI: at one tech company, the lawyers already use an internal tool to generate contracts. “Their job isn’t really to write any more, it’s to proofread or revise,” we’re told. In so doing, meta-employees complete the work of the AI (which is also their own): “in our two- or three-month sales forecasts, we realised that an algorithm was generally better than a human being; but apart from that, planning experts are always more accurate,” says Pierre-Louis Bescond (Roquette). “It’s a bit like an oncologist using AI to detect abnormalities: the algorithm achieves 95%, but when you put the two together, you get to 99%.”

A PILOT

A pilot also implies a copilot. Microsoft has picked the right name for its chatbot: it’s THE metaphor most frequently used by the managers we interviewed. “AI is a copilot, but the pilot is in charge of the aircraft,” was something we often heard. Meta-employees are not seen as employees who would let go of the joystick. “They are the ultimate managers,” says one HR player. “If they use Generative AI to develop a product, they remain responsible for that product.”

A CREATOR

Meta-employees make AI in their own image. As part of the organisation’s change management process, they are fully involved in the development of AI projects that concern and can influence them. At Malakoff Humanis, employees in the Fraud Department “are involved in creating a project: we need their expertise – they are the driving force,” says the Data Product Manager, Jean-Luc Leblond. PierreLouis Bescond goes even further: he imagines a personal AI he has trained himself, which is “used to [his] style and way of working.” The meta-employee is thus a creator who needs to be creative: “We can’t ask AI to create a finished product,” smiles one digital specialist. “We need highly imaginative talents” to manage it. Meta-employees go beyond what already exists, and develop another reality: their creativity is expressed in their prompts.

A PERPETUAL STUDENT

“I read an article where an expert described how a virtual assistant had been invited to a meeting – though nobody knew who’d done it. As soon as the meeting was over, it sent a transcript to all the participants, without proofreading or checking it. The emergence of these tools can create risks,” says Pierre-Louis Bescond. Meta-employees are trained in the uses and potential dangers of AI. “It means taking a responsible stance,” says Eneric Lopez. “Being able to say ‘I need to learn’.” Continuous learning is needed to keep pace with developments in AI.

A PRAGMATIST

The efficient meta-employee accepts AI. They “desire” it because it can remove “the stone in their shoe”, to quote Eneric Lopez, director of Microsoft’s National Initiative for AI. “It’s important not to have a techno-centric approach to AI but a business value-centric one, and to ask yourself how this tool can save you time.”

A DEPENDENT EMPLOYEE?

You soon get used to good things. When the in-house AI solution went down for a few hours, PierreLouis Bescond, “for the first time”, saw Roquette employees arrive on the Data team’s platform. How is it possible to do without a tool you’d never even used a few months ago? The dependence AI can create is not the only pitfall awaiting the unwary meta-employee: our interviewees cited a loss of vigilance and the temptation to do less. This is the dream of otium: “the highest way of living for the Greeks and Romans,” says philosophy researcher Luca Paltrinieri. “For them, fulfilment was achieved not through work, but through the time that was freed up.”

A COOL-HEADED EMPLOYEE?

Eliminating bias. This is what we expect from AI, particularly in human resources. Doing away with prejudice, feelings, angst and all the subconscious factors that influence our actions – and assuming post-human decision-making.

A MANAGER

“When I look at the tasks replaced by AI, I tell myself that resistance is impossible: AI will replace me one day. But work is not a task. Work is not even the sum of several tasks. Work requires something extra: the ability to coordinate several tasks. And that’s something artificial intelligence can’t do at the moment,” says Luca Paltrinieri. “As a researcher, I write, read and teach. This coordination of different activities is a form of management.”

A COMPETING WORKER?

Imagine being in competition with your own colleague! We can imagine some people picturing it perfectly well… For the meta-employee, this is also true. “AI is a new form of competition,” says Luca Paltrinieri. The English queen Elizabeth I had the inventor of the knitting machine’s assistant executed for fear that it would destroy jobs. The key is to turn it into a tool, not a machine. The difference between the two is human knowledge.”