Pascal Montagnon, Artificial Intelligence Specialist

On the occasion of the Artificial Intelligence (AI) Summit held in Lyon, Lyon spoke with an artificial intelligence expert at Capital.

Pascal Montagnon, Artificial Intelligence Specialist
Pascal Montagnon @ Omnes Education

Pascal Montagon is the director of digital research, data science and artificial intelligence at OMNES Education Group. As such, he is also the organizer of the 4th edition of the Artificial Intelligence Summit which is being held in Lyon on Tuesday 4th June. On this occasion, Leon Capital met with him to discuss artificial intelligence and its impact on the evolution of today’s society.

Leon Capital: You are the director of a chair of digital research, data science and artificial intelligence. Can you define these three terms for us?

Pascal Montagon: These three words are complementary, because we can’t talk about artificial intelligence without talking about data. Artificial intelligence exists only through the value of the data you collect. Today, it is important to understand that data is everywhere. When you consult the Internet, for example, you leave a trace. This will provide trace information. This data will be used in a variety of ways to better understand who you are: your eating habits, your outlook on business, and so on. In most cases, when you click on a website, you are asked if you want to allow this data collection, which 99% of people do. Therefore, provide material to those of you who will use this data. This date is called. One of the purposes of artificial intelligence is to be able to make predictions using these data. And the more data there is, the more predictable it should be – Descending – To be reliable

Many companies talk about artificial intelligence but in reality they don’t …

The starting point is to make a diagnosis: “What do I need for my business? What kind of information do I need to be able to improve my performance?”. Diagnosis is essential for achieving results that are consistent with needs and limit investment. Everyone says that artificial intelligence is very expensive. No, today we can create a very simple algorithm according to your request.

Credit: Pixel.  Robotic hand of artificial intelligence
8 pixels

Artificial intelligence is a matter of imagination, especially staging, Hollywood, A competition with human intelligence. What are the similarities and differences between these two intelligences?

If the word intelligence is similar to theirs, it stops there. I’m not part of this school that thinks we’re moving towards transhumanism, meaning that tomorrow robots will replace humans. While it is not possible to talk about artificial intelligence, robots have been around for decades. When we talk about automation Process, There are actually robots already, but in no way does artificial intelligence replace human intelligence. First, because we need people who create algorithms. Then, because the algorithm should be built around something that can be improved, and only humans can see the biases and correct them.

“We can’t understand feelings or ask for an algorithm to have emotions. It’s impossible.”

Third and last point, we know that an algorithm is a mathematical “machine” that is foreign to any concept of feeling. No.We cannot ask an algorithm to have feelings perception or emotion, it is impossible. Let me give you an example: When hiring, some companies use software. In other words, the candidate’s CV is analyzed by an artificial intelligence based on keywords. When a video is requested from a candidate, it is still artificial intelligence that, for example, through face recognition, will select this or that candidate. It is only in the third stage that people, i.e. employers, intervene. But artificial intelligence cannot do everything. When I was a director of a company, I had to deal with candidates who did not stick to the position they wanted but on the other hand, had qualities for other positions. I put them in this post and it stuck. Artificial intelligence does not know how to do this because it is programmed to recruit a person for such and such position with such and such features.

Can you give us examples of everyday life where artificial intelligence exists, if not omnipresent?

In two or three years, bank cards will gradually disappear for payments on our smartphones, a real focal point for artificial intelligence, a digital application that will be based on artificial intelligence algorithms.

The second example is autonomous vehicles. If they exist, we still can’t keep them on the road because there are still many devices that we don’t know how to handle. For example, in case of an accident with an automatic vehicle, that is, without a driver, who is responsible? Is it the car maker or the people inside, or the pedestrians passing by? There is a legal vacuum in the matter. An autonomous vehicle can run today when there is no hassle around. If you leave an autonomous vehicle at 6 pm on the Place de l’Etole in Paris, it will be completely disrupted. This is an improvement that science is working on.

8 pixels

The third example is the perfect, video surveillance camera in cities. After the attack, Nice Significantly furnished promenade des angles. This makes it possible to use facial recognition, which means that a person whom the police want will be automatically detected by the camera, which will allow the police force to do their job.

The Artificial Intelligence Summit, held in Lyon, sought to link economic questions with research around artificial intelligence. What will artificial intelligence bring to the economy and management?

In the economic world, there is a lot of human intervention in processes. But who says human intervention is the source of error. This will make the artificial intelligence processes more reliable, which will become more efficient if the algorithms are well developed. In addition, artificial intelligence brings speed: an algorithm will count to one millionths of a second that you and I could count in five minutes. The speed with which the death penalty is carried out will bring an exploitative outcome that a basic human being could not produce. We see this in our daily lives.

“For management, artificial intelligence raises a question that entrepreneurs cannot deny.”

For management, artificial intelligence raises a question that entrepreneurs cannot deny. Along with new skills, new professions also appear. For example, data scientists who appeared a few years ago did not have to create other professions. We also train them in our school, because it is important for managers to be able to understand the algorithm, because they cannot rely entirely on a data scientist who can manipulate the algorithm to create bias.

8 pixels

Let’s take a well-known example: Amazon had to review its hiring algorithms because they realized that the algorithm preferred men over women. When these biases are corrected quickly, they do not have much effect. The same thing happened with Amazon. In management, new technical skills and new professions need to be managed. After all, today a manager sees his legitimacy in that he is a technical reference in his field. However, with artificial intelligence, it will no longer be necessary. Thus, instead of managing difficult skills such as technical skills, managers can focus on soft skills consistent with behavioral aspects.

“Today, a system that wants to keep your team around you is outdated.”

It has not escaped anyone, including the epidemic, telework has appeared, or at least it has developed strongly. To guarantee their attractiveness, companies must offer permanent face-to-face options. We also find artificial intelligence behind this question of telework. Today, a system that wants to have your team around you is outdated. Managers need to learn to manage other forms of work.

What are the challenges facing artificial intelligence?

I don’t do science fiction, although others do it their business. In my opinion, artificial intelligence is a technological development whose applications will enable it to survive better, at least separately. At present artificial intelligence is not capable of doing much. It was recently pointed out to me that artificial intelligence was not able to predict the Covid-19. This is explained by the fact that, since the situation was unprecedented, there was no data on the subject. In my opinion, the main challenges of tomorrow’s artificial intelligence are to keep up with the power of the tools and especially the so-called quantum.

“Quantum will give us access to more powerful and faster computers.”

For example, our current smartphones have the power of computers used during the 1992 golf war. For some time, technology has been advancing at a significant rate, but in recent years we do not know how to create more powerful products. However, Quantum will bring a new phase of progress and we will see super powerful computers including supercomputers. With 5G, we will use more data and without Quantum we will not be able to process this data. Quantum will give us the opportunity to get a more powerful and faster computer. We have no choice: we have to rush to 5G because when we are debating, the Chinese are already working on 6G and there is already a war ahead of us.

Leave a Comment