Can we still do without artificial intelligence in IT?
Artificial intelligence (or AI) is when a solution is capable of adapting to an environment in order to learn, but also to understand and think logically. To achieve this, such a solution needs to process large amounts of data and must be able to make decisions itself.
Of course, all this is no longer science fiction, but apart from a few examples such as personal assistants or autonomous vehicles, the development of AI may still seem limited. However, this technology is already appearing in several sectors of activity. Whether it involves automated processes, decision support or customer relations, projects are multiplying. Soon, AI will be unavoidable.
Although users still express some reluctance towards AI, companies seem to have seized the opportunity and even the need to integrate this new technology into their IT strategy. For some large groups such as Thales, AI is already being put into practice. And this concerns all trades.
Indeed, AI offers a number of advantages in terms of optimization: for example, it can detect anomalies in the production chain, or even recognize objects and people. It remains for companies, especially CIOs, to grasp the areas essential to AI, such as machine learning, deep learning and data analysis.
To this, we must add what is called symbolic AI, i.e. constrained programming or so-called genetic algorithms, because the stakes are high. Mastery of these different technologies should enable professionals, in the long term, to solve a certain number of complex problems and help them to make automatic decisions.
If we simply take the example of health, the perspectives are indeed broad: predictive medicine, companion robots, computer-assisted surgery, etc. The same applies to the management of energy consumption (via the development of smart grids that make it possible to anticipate the production and distribution of green energy at peak times), or the exploitation of natural resources.
Sectors already adept at AI
Telecoms and services are also particularly attentive to the development of AI. It must be said that it can represent for them a fantastic commercial advantage, especially if it is in direct contact with the end customer (via chatbots or personal assistants such as Amazon Alexa).
It can also be integrated within applications, such as the YES SNCF application that is able to predict your next destination, but for the moment, AI is mainly used to optimise the performance of promotional - or loyalty - campaigns via predictive scoring procedures. One of the companies that have decided to take the plunge is Bouygues Télécom.
The operator has signed a contract with the start-up Notify to improve the efficiency of its call centers thanks to "Best-Time to" technology (which identifies the best times to contact customers). Conclusive tests have shown a 39% increase in outbound call reachability. For its part, Softbank is a reference in terms of AI.
The telecom group, after having bought Frenchman Aldebaran, caused a sensation last year when it announced the acquisition of Schaft and especially Boston Dynamics, an American start-up known for its humanoid robots. Another sign that does not deceive about the future: although it innovates little at the moment, SoftBank plans to become a major player in AI.
The Japanese group has even signed a partnership with Saudi Arabia to create the largest technology investment fund in the world.
Artificial intelligence rhymes with computer security
Much more wary, private banks can also benefit from the advantages of AI, especially for their asset management. According to specialists, this technology could have an impact on all value chains: sales and marketing optimization, real-time customer relations, predictive market modeling, and why not robots to process data and help perform financial analysis... But AI could also provide banks with serious guarantees in terms of IT security.
In this respect, we note a reluctance to use the blockchain, which is itself the origin of Bitcoin. However, in this area too, interesting initiatives are emerging. The company Neurochain intends to go beyond the concept with more intelligent algorithms (to precisely measure the quality of exchanges and interactions, user perception, etc.).
In this way, its decentralized application allows professionals to work with confidence. And AI obviously plays a major role in detecting fraud or anomalies in the block chain more quickly.
In this respect and well beyond the banks, companies have understood the interest of AI to fight against computer attacks (knowing that they themselves are increasingly using AI). According to an Alfresco study, 21% of IT decision makers believe that AI is now mature enough to help them protect their business.
Another interesting piece of information: 99% of CIOs say they are ready for the AI to take security measures alone and almost one in ten (8%) already do so via the learning machine. In some structures, the learning machine replaces tedious tasks and supports decision-making. Advantage: IT teams can turn to tasks with higher added value.
One thing is sure: the revolution is underway and for IT services, it will soon be difficult to do without AI to ensure competitiveness and security. The only thing left to do now is to raise the persistent concerns. To do this, public authorities must take the necessary measures and think, with the stakeholders concerned, about an effective framework for AI, as well as all the new technologies that can be associated with it.