Robots, implants, sensors, automation, 3D... The field of artificial intelligence (AI) is vast and is upending our work and private lives. But what is it really changing? What are the limits? Cédric Vasseur, a lecturer and expert in artificial intelligence, offers AXA Partners an update.

When it appeared on stage with its strange metallic structure, Elon Musk’s humanoid robot did not give the impression that it was an exceptional development. However, the co-founder of SpaceX asserted that his robot would set in motion “a transformation of civilisation”. Named Optimus and expected to be available in three to five years, the prototype uses artificial intelligence to move around and pick up objects independently. The goal is that it will eventually relieve human beings of the need to perform repetitive tasks.  

After all, that is one of the key objectives of AI. From the comfort of our living rooms to the doctor’s surgery, not to mention our workplaces, AI is revolutionising the way we do things, both at home and in the office, explains Cédric Vasseur, a lecturer and expert in artificial intelligence. 

Preventing and evaluating accidents

Everyday benefits include an ability, enabled by machine learning, a subcategory of artificial intelligence, to predict and thus protect us from household and industrial accidents – for example by sounding the alarm if the oven is left on too long, or detecting an appliance that is overheating or an anomaly in a system. 

But how does this work, exactly? Using sensors that transmit details of vibrations, humidity and temperature, the machine learning algorithm analyses the data and diagnoses faults at an early stage. If an incident occurs, robots are in the front line. “Their assistance can be decisive in high-risk situations. Remember that when the fire occurred at Notre-Dame Cathedral in Paris, it was Colossus, the firefighters’ robot, that was the first to enter the location,” notes Cédric Vasseur.  

All by itself, Colossus extinguished part of the fire and reduced the temperature inside the nave. This technology, which can be deployed on aeroplanes or in self-driving cars for example, will help to better prevent some tragedies. This assistance is particularly useful in areas that are inaccessible to humans, where robots can get in, rapidly analyse large volumes of information and direct rescue teams.  

Supporting patients and changing lives

One of the most promising sectors is health. One example is patients affected by hearing disorders. “Some people who are hard of hearing can now hear using artificial intelligence,” said Cédric Vasseur. This progress has been made thanks to cochlear implants equipped with a processor that captures sounds and transforms them into electrical impulses that activate the implanted stimulator, making it possible to follow a conversation. “France is even at the forefront on this issue, and offers full reimbursements despite the very high cost,” explains the AI expert. 

Robots also have an important role to play in managing dependency and old age. “Many elderly people do not wear their alert bracelets, but if they have a voice assistant in their living room, they will be able to quite naturally ask for help or cry out if they fall,” Cédric Vasseur points out. Now widespread, these voice assistants can also help to combat digital illiteracy, allowing older people to send a text message or dictate an email.  

2 people watching an electronic device

Analysing emotions

Over the last few years, some customer service departments have been using conversational robots, who respond to some customer queries in writing via chatbots and also by telephone, with assistance from a synthetic voice. “We, as human beings, are not always great readers of emotion,” explains Cédric Vasseur. “AI makes it possible to detect feelings more accurately and guide customers accordingly. It can even analyse the key words that will make you react and adapt. This means that it is possible to train fewer telephone operators. Previously, we used to write out scenarios for them on paper! Now, what to say is displayed on screens, they just need to click,” smiles the expert.   

Some emergency services in France use this technology to detect heart attacks more quickly and increase patients’ chances of survival. Real-time predictive speech analysis monitors conversations to detect signals such as voice intonation, respiration rate, background noise and the words said by the person in distress.  

Assistance and insurance: when AI is used to support people

In the fields of insurance and assistance, AI can facilitate the work of staff by enabling the automation of some repetitive tasks and supplying accurate and up-to-date information in real time. In this regard, AI can help insurers to evaluate risk and price policies more effectively. The collection of data and partial automation of these risk evaluation and pricing tasks also make it quicker to take out an insurance policy, something which helps both insurance agents and people looking for insurance, and to evaluate risk more accurately, meaning that prices are better aligned with customers’ actual needs.  

With respect to the customer experience, AI-based chatbots can be used to help insurance policyholders to find out information about their insurance policies, report incidents or obtain assistance when required. Some tasks can then be “delegated” to an integrated AI, which will gather certain information from the customer, verify their request (using image recognition among other things) and offer a range of solutions.  

In this context, progress made in natural language processing (NLP) in recent years has enabled the use of virtual assistants. If an accident or a breakdown occurs, for example, chatbots can be used in addition to insurance agents to help customers receive a rapid response. Customers can report a breakdown, call a mechanic and even choose a replace vehicle directly from an API or a mobile app that includes a virtual assistant. These solutions enable better handling of requests for assistance and at the same time provide relief for claims management departments. In addition, these chatbots can be accessible 24/7, enabling policyholders to receive assistance at any time.  

Thanks to its speed and precision, artificial intelligence can predict certain dangers more accurately and assist accident victims under any circumstances.  An opportunity which has already resulted in a tangible development for the insurance sector, which, in addition to its industry expertise, can continue to offer the best possible service to its customers.  

More inspiration?