• Erwan Hernot

Will AI Be Used For Human Talent Optimization?

Most experts have made great efforts to distinguish what artificial intelligence can do and what it cannot. The differences, based on strengths of both humans and AI are pretty clear. But, to me, the real question is not this Human/AI classification but rather another one : will executives use AI for human talent optimization and consequently for a better customer experience or will they stick to short term financial goals (and an acceptable-but-not-so-great customer offer) ?


First, as a reminder, the usual classification : what AI can do and what Humans are better at.

What AI can do:

  • Coordination of work and planning, scheduling . For example: setting a schedule based on consumer demand,

  • Record keeping. For example: meetings and their follow up,

  • Data mining and gaining insight through extensive analysis of structured data. For example: some of the insights AI generates will be fed to human decision makers like a recommendation to a sales person about which customer to call on next, most often using machine learning,

  • Quality control, performing audit for compliance with standards,

  • In HR, controlling over the employees. For example: an internal operating system that tracks things like billable hours, workflow and employee attendance. AI, with its unbiased and ruthless logic, could prove a leveler in the selection and recruitment of candidates and in ongoing performance reviews and remuneration policies.

  • And, eventually, problem solving and logical reasoning when problems are known and data is available.

Advancements are coming too, in capabilities including the ability to deliver messages with nuanced human interaction. Engaging with customers and employees using natural language processing chalkboards, intelligent agent and machine learning.


What Humans are better at:

Humans possess a very general mental capability that involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. Today, artificial intelligence is not intelligent in these sense. Machine learning does only the last two items in the list: learn quickly and learn from experience. AI can indeed recognize patterns in data and then act or make suggestions based on the patterns they find. This patterns recognition is super human in specific areas. But AI is below the average human performance for creativity, identifying new patterns (no data), logical reasoning and complex, ill-defined, non immediate problem-solving. Humans can innovate and develop thoughts and notions that are not based directly on past experience. Humans can plan for things that might happen and solve problems at a general level without nailing down all the details. Big-picture thinking is one capability at which managers are still better than computers—and will continue to be for some time. Machines are not very good at piecing together a big picture in the first place, or at noticing when the landscape has changed in some fundamental way. Good human strategists do this every day. That includes thinking of novel ideas to solve systemic issues. Managers understand and respond to others' emotions, pull and build raw talent, and empathize and support people. AI is not socially aware or able of empathising either, thus doesn’t sustain relationships that businesses need.


There are decisions to be made, what could they be?


Executives face the classic trade-off Exploration vs Operations. Operations are based on the efficient use of acquired skills, it produces reliable results but presents the risk of missing out on more promising alternatives. Exploration consists in looking for new possibilities at the risk of not deepening them enough to reap the benefits of their mastery. In a situation of failure, exploration tends to prevail. Conversely, a successful operational model tends to discourage exploration whose benefits are uncertain, while the experience gained in incremental and “successful” adaptations of the Taylorist model currently used makes operations more attractive. Two reasons tend to make me pessimistic about executives’ decisions about AI augmenting new managers (aka learning leaders) : an error of judgement and a preference for short terms profits.


Error of judgement


As far as we still live in a Taylorist era, managers stick with their traditional roles. Some experts quickly point at theirs weaknesses: managers are top down information transmitters, furthermore imperfect ones : ”Managers are humans, which implies that by all means they perceive and interpret information subjectively. Also, some would add their own opinion and vision. All in all, the initial message would not be delivered unchanged. Their primary role is to convey the strategy to lower managers – the task which can easily be solved by messengers, video conferences etc. Information flow within a company should be fast, and today’s technology make this possible. And with the advent of smart technologies, the need in such professionals is shrinking.” This assertion comes - of course - from an AI solutions vendor. Beyond information transmission, you have decision making and we hear the same music: “machine learning is making possible more efficient, effective and optimized business decisions. Given the unreliability of human decision makers and the difficulty of interbreeding machine learning, it’s likely that more and more decisions will be made and even executed by machines over time. Some of those will be typical decisions made by middle managers such as what person to assign to a particular project”. RPA (1) aims at automating all the possible processes and tries to suppress middle-level managers along the way. Overall, across industries, frontline managers spend only 10 to 40 percent of their time actively supervising their employees, according to McKinsey report. At some companies, managers devote just 4 to 10 percent of their time — as little as 10 minutes a day — to coaching teams. In these 2 examples, everybody thinks with the current picture. No one envisions a double change: AI is coming AND a new breed of managers (the Learning Leaders) are emerging.

Short term profits


Indeed we live in a world where corporations are biased toward the narrow interest of theirs shareholders. Years ago, companies’ management have shifted from focus on long-term growth to getting their stock price at a certain level. The paramount concern with the foreseen immediate consequences excludes consideration of further consequences. We can’t, then, be optimist on the way AI would be used. How many chief financial officers will be willing to forgo investment in training managers to their new role and plan to upskill employees? AI/Human partnership is a project expected to be profitable in the long term, not able to meet analysts’ quarterly earnings estimates. Furthermore, executives could just rely on AI to increase productivity. Cutting costs and “making more with less” is a mantra known from the eighties. Why should executives “programmed” to manage rather than innovate would charge AI with this bold mission of increasing human value by working with managers? Chances are that they will replicate what they know because they just understand cognitive technologies as a tool to save money.


AI as human augmentation


“Executives must develop a clear thesis on the future of work. Artificial intelligence must be used to make current workers more successful, not to lay them off.” said Paul R. Daugherty (2). Will he be heard? If companies want to stay ahead of competition, executives will use cognitive technologies to augment managers rather than automating their job completely. Indeed, many decisions require insight beyond what artificial intelligence can squeeze from data alone. Good managers will use their knowledge of organizational history and culture, as well as empathy and ethical reflection. This is the essence of human judgment — the application of experience and expertise to critical business decisions and practices. Managers have already a sense of a shift in this direction and identify the judgment-oriented skills of creative thinking and experimentation, data analysis and interpretation as the top new skills that will be required to succeed in the future. It will increasingly be the manager’s responsibility to lead the company and its employees through the transition as seamlessly as possible to humanize the process and make sure the company is set up for optimum business results. Taking a talent optimization approach to the integration of AI in the workplace will allow managers to effectively redesign their teams and ultimately to skillfully redirect their time toward more human activities such as increasing customer satisfaction. If companies focus on the solely goal of saving money, they will induce bad management that will only use AI as an automation agent and misuse it as a (poor and basic) problem solver.

(1) RPA stands for Robotic process automation is a form of business process automation technology based on AI.

(2) Paul R. Daugherty Human + Machine: Reimagining Work In The Age Of AI, 2018


  • Noir LinkedIn Icône
  • Noir Twitter Icon