Cosmo Tech’s Executive Chairman and co-founder Michel Morvan recently gave an interview for Paris-Saclay about his involvement with IRT System X and the future and limitations of artificial intelligence.
In April, Morvan was unanimously voted as the President of IRT SystemX, after being actively involved in the institute since it’s conception in 2012. IRT SystemX is an institute for technological research, aimed at strengthening territorial attractiveness while boosting the digital transformation and representing a new dynamic of innovation for France. At their first event taking place on 22 May, Digihall Day, Morvan gave an inspiring call to arms for French research, small businesses, and corporate worlds in an era of digital transformation. His interview was centered around the French innovation ecosystem and a topic Morvan is particularly passionate about – artificial intelligence.
Morvan, a household name in the world of artificial and augmented intelligence, understands the characteristics of AI that make it so powerful. “The purpose of AI, if there is one, is to copy human intelligence,” he says. “However, this presents two characteristics schematically. The first rests on the recognition capacity of our brains.”
“The other characteristic of human intelligence,” he continues, “resides in the abilities to understand and to reason. For the most part, current AI consists of nothing more or almost nothing more than recognition. Not comprehension. If, today, we are able to teach machines how to recognize, we are not yet able to teach them how to understand. It will take several more decades before we get there. If we ever do.”
However, Morvan also understands the challenges that AI poses. “AI, as a recognition tool, is essentially based on data from the past. But everyone admits: the world is constantly changing and evolving in directions that are hard to predict! [AI] can only help to predict events that have already occurred, certainly not to anticipate events that have not yet occurred i.e., ruptures or so-called “black swans,” he states. “However, for a senior executive, the important thing, is being able to predict and anticipate that which has not already happened.”
While the computational processing that AI can produce is certainly impressive, Morvan understands the importance of the analytical capabilities of industry experts and how they are indispensable during the decision-making process. The ability to understand and update causal relationships as opposed to just correlations is another limit of AI. This again falls under the category of AI’s inability to understand and reason, which is admittedly its biggest constraint. For example, when playing a game, AI doesn’t rely on an understanding of the game, however instead relies on the systematic and methodological recognition of patterns and positions played.
Morvan has spent significant time in Silicon Valley, most recently returning from a 3-year stint there. While Silicon Valley is especially known for their technological advances, Morvan was optimistic on whether or not France was lagging behind.
“I am convinced…that France can become the world leader in the digital transformation of industry and its systemic issues,” he says. “Far be it from me to say that France can easily become as great as Silicon Valley in B to C by creating the next Google and Facebook. However, [France] can establish itself in the digital transformation of industry, otherwise called Deep Tech B to B”. Continuing on the topic, he says, “My hope is that tomorrow, we will stop being surprised at the possibility of this dialogue between academics, industrialists, and entrepreneurs, that it will develop to the point of becoming common place. However, such a dialogue in Silicon Valley has long since come as no surprise.”
Read the original interview here (in French).