THE CHALLENGES OF TEACHING AI

Date

November 2, 2023

Description

Laurent Astoul, director of the IA Institut, talks about the challenges of teaching AI.

image

Laurent Astoul's biography

Laurent Astoul is Director of the IA Institut.

Laurent Astoul arrived at the time of the data acceleration in the 2000s, and has experienced all the technological breakthroughs in this field. After eight years of expertise acquired in consulting firms, he implemented delivery and out sourcing rationalization strategies on an industrial scale at a major telecoms operator, and designed data solutions for all the company's businesses. Two years dedicated to the commercial development of a pole of expertise at an integrator strengthened his conviction that career development and quality recruitment play a key role. His knowledge of business models was further enhanced by four years in charge of the data departments of a pharmaceutical laboratory and then of a major building distributor. Convinced that training and interpersonal skills are the real keys to success, Laurent Astoul chose to develop the business of the IA Institut school, positioned on the future technologies of artificial intelligence and data engineering.‍

OPTIC: Laurent Astoul, you're Director of IA Institut, the data and Artificial Intelligence school. In France, how do you see the education system adopting AI?

Laurent Astoul : Waves of growing needs in AI follow one another. The second phase, dated 2021, of the PlanFrance 2030, led by the Agence Nationale de la Recherche, illustrates the current historic moment of regrouping and consolidating several French centers of excellence in Artificial Intelligence training. The aim is to build a network of five to ten French academic clusters, and encourage the emergence of two to three internationally renowned clusters of excellence. This is a national and European strategy. President Macron recently announced an additional €500 million for AI at the VivaTech trade show. France is undeniably positioning itself.

OPTIC: Do you have any idea of the current need for AI training in France?

Laurent Astoul: Things are moving very fast! We can mention the July 20 study by Le Monde Informatique, which indicates a 600% increase, between May 2022 and May 2023, in recruitment ads mentioning "ChatGPT" or "GPT". According to Upwork, in 2023, the number of job ads referring to generative AI skills such as ChatGPT,Bard, etc., jumped by 450% compared to 2022. It would seem that we are almost at the peak of expectations for AI technologies. On the other hand, let's be cautious about the potential disillusionment phase, should current returns on investment be less fruitful than expected.

OPTIC: How is it possible to meet such a high level of AI training needs in such a short space of time?

Laurent Astoul: If it is not possible to train so many people in such a short space of time, it is possible to optimize the response to needs. The private education system is very reactive and proactive. Once the time needed to implement new training courses has passed, the whole system will accelerate rapidly in the field of initial training. Ongoing training for people working in organizations is a powerful lever for transformation, provided that the necessary investment is made. We can rely on French players in the sector to help with professional retraining, in this case for AI training in the current profession, within a relatively short timeframe of around one to two years.

OPTIC: In your opinion, are French organizations and companies positioned to trigger their transformation thanks to AI?

Laurent Astoul : In my opinion, yes, because all the companies and organizations I work with are in the process of transformation. They need to be supported in this evolution. Over the past year, we have seen a succession of trade fairs and inter-professional conferences about AI. In addition to the war for talent, there is even currently a war for AI speakers, and we are very much in demand. What professionals and decision-makers are telling us right now is that AI has been on the agenda of many executive committees for the past two or three years. With the international buzz of ChatGPT in 2022, executive committees are now demanding action plans for AI-enabled transformation.

OPTIC: Precisely, do you think that the executive committees of organizations find their way around the AI vocabulary, that they distinguish between generative AI and other AI systems?

Laurent Astoul: Absolutely not! It is not uncommon for a member of an executive committee to be less experienced in ChatGPT than his or her child in high school! Admittedly, the general public still has a hard time grasping this "science fiction"scenario, in which a machine advises a human being. Do you remember HAL inStanley Kubrick's 2001: A Space Odyssey? Would you bet a penny on the likelihood of this happening two decades later? I doubt it. All that is left is to give a voice to generative AIs like Bard or ChatGPT, and we will be there! Overall, human acceptance of being surpassed in finesse and intelligence has not yet arrived. It will arrive the day AI performs a great service for humanity. Only then will AI be fully accepted and embraced.

OPTIC: IA Institut is a French player in the field of AI education. What conclusions have you drawn for your own training offer?

Laurent Astoul : We need to respond to two key training challenges at the same time.Firstly, we need to put in place, without delay, the initial training courses we need to supply the market with skills in the years to come. Secondly, and what I am telling you is far from anecdotal, the real transformation will involve training existing staff in organizations, the millions of people whose jobs will evolve with the contribution of AI or who we will be asking to contribute to projects to transform organizations with AI. It is a question of continuing training, derived from initial training, adapted for the profiles that will require it. This is how we orientate our initial and continuing training offers.

I would like to add a final, important and cross-cutting key, which is the quality of training, which must integrate the fact that AI is not a technology, it is a new discipline and, like any discipline, such as medicine, it can be applied in a wide variety of situations. To have a good training in AI, in addition to the technology, it is the whole mindset that goes with it, and that is key: ethics, data regulation, awareness of the impact on job transformation, knowledge of the profession for which AI is used. In other words, in addition to teaching technology, it is essential to teach the jobs, particularly those taught in business schools. At IA Institut, we take a 360° approach to teaching AI, combining the fundamentals of business with science and technology.

‍