BusinessToday Previous Editions

23.01.2020 8 OPINION Alexiei Dingli Prof Alexiei Dingli is a Professor of AI at the University of Malta and has been conducting research and working in the field of AI for more than two decades, assisting different companies to implement AI solutions. He forms part of the Malta.AI task-force, set up by the Maltese government, aimed at making Malta one of the top AI countries in the world B efore dawn breaks over the city, John wakes up to go to work. It is an important day for him be- cause he has to give a presentation to the board of directors. Since he was a bundle of nerves and the day appeared to be rath- er pleasant, he decided to drive to work rather than allow the self-driving car to take him. John had subscribed for a smart in- surance plan through his mobile device. It essentially means that the insurance rate varies based upon his circumstanc- es. e fact that he decided to switch off the autopilot of his car resulted in a slight increase of his insurance pre- mium for the drive which would go up by a few cents. But that is fine for him because he needs to distract his mind from the presentation. Every year, 1.25 million people used to die worldwide due to car accidents with a further two million people in- jured. 94% of these crashes were due to human error, but with the proliferation of self-driving cars having more than 60 years of driving experience each, the number of deaths and injuries shot down. e increase in safety eventually resulted in a reduction in insurance pre- miums globally. As soon as John got comfortable in the driver's seat, his virtual assistant Katy checked his diary and programmed the car's navigator. is route not only took into consideration the traffic at that hour plus the driving time, which would allow him to arrive at his destination in time, but it also considered safety con- siderations. Since all cars have a live connection to a centralised Intelligent Transportation System (ITS), the Arti- ficial Intelligence (AI) harvests the data of millions of vehicles every day. Essen- tial statistics are then calculated, such as the traffic and accident hotspots, the likelihood of accidents, etc. e system optimises the entire road network to re- duce traffic and accidents. However, John decides against this route and chooses one which is more scenic but which is not recommended by the system. is choice results in a further increase in his daily insurance premium because the risks are slight- ly higher. ese increases get charged through his bank account automatically using a system of micro-payments. Along the way, a crazy driver still going around with one of those vehicles (with- out crashing sensors) bumps straight into him. e car has an internal di- agnostics system which automatically determines the extent of the damage suffered. e statistical analysis which John's virtual assistant Katy ran on the accident categorised it as being minor, so no ambulance was dispatched. How- ever, it interacted with John to ensure that he was feeling fine. Furthermore, realising that John was going to be late for his presentation, Katy summoned a replacement vehicle to take him to his office as quickly as possible. In the meantime, the ITS sent forth the traffic drones to document the ac- cident with video footage. Katy had already raised a claim with the online insurance company and uploaded all the documented evidence. In this case, the accident was quite straightforward, and the AI judge ascertained the fault of the other human driver. Since the car was still operational, no self-driving tow-trucks were required, and the car drove itself to a nearby panel beater for repairs. All the expenses incurred were charged to the other human driver au- tomatically according to the decision taken by the AI judge. When the pan- el beater finished the repairs, the car drove itself back to John's house. In the meantime, Katy kept John updated with the status of the repairs. Of course, this is a futuristic scenar- io, but it might be much closer than we think. Self-driving cars are already available and used extensively in some states of the US. Virtual Assistants, too, can be found in almost any mobile phone, from SIRI to Cortana. ese assistants already help out in the navigation, and take into account considerations such as live traf- fic, appointments, etc. Many countries have an ITS in place with an army of video cameras and oth- er sensors capable of monitoring road networks and dispatching assistance when necessary. Drones are becoming popular and ac- cessible. ey are already very success- ful in surveillance of remote areas and delivery of light goods. AI judges are already taking minor de- cisions. As an example, Estonia is cur- rently piloting a "robot judge" that can adjudicate small claims disputes. e different components are availa- ble. What is still lacking is the glue to link them together. However, it is only a matter of time before we start living the scenario mentioned above. When we do, our lives will change forever. We will have fewer car accidents and few- er fatalities. We will be able to concen- trate on what matters most and not get worked up with driving stress. For the boring parts, there's always our virtual assistant who is working tirelessly to make our life a breeze. Killing the motor underwriter Every year, 1.25 million people used to die worldwide due to car accidents with a further two million people injured. 94% of these crashes were due to human error With the proliferation of self- driving cars having more than 60 years of driving experience each, the number of deaths and injuries shot down

• Cover