AI between challenges and opportunities
Interview with Fabien Mangeant
AI is a digital technology that has been developed to mimic some human behaviors. So it's about knowledge search, pattern matching, decision making. And this technology that we want to use to assist humans is used already in many areas. So this is used in healthcare, to develop diagnosis. This is used in mobility to support road traffic prediction. This is used in weather forecast. And it's already bringing a lot of benefits.
AI has been developed inside Air Liquide for more than 20 years. This has already been used in production planning, production scheduling support to the supply chain, minimizing the number of kilometers driven by our trucks. This is also installed inside our customers to understand and predict their demand. We are also using AI for home healthcare. We use the data collected from the healthcare devices and give them to the nurse, the technician, at the right moment.
As a rising technology, AI is also bringing a lot of new challenges. We know that the datasets that are used to develop some of the AI may be biased. So this means that we could have a bad influence of these AI on the global behaviors of the society. We know also that there may be lack of transparency, lack of explainability.
To make sure that we use AI in the right way, we have to develop a lot of new technological bricks, to develop the governance and ensure that we are using a trustworthy AI.
At Air Liquide, we strongly believe that AI will help our employees with new tools, so they will have access to new informations, to new knowledge, to the best practices of the company. And we will also be able to automate the most repetitive tasks that they face every day. This is going to be really the revolution of the years to come. From a societal point of view, I wish that AI will help us solve the most complex problems of our world. And I wish also that of course, we will use a trustworthy AI with a minimized carbon footprint for the good of the society.
