11 JUL 2018

5 examples of applied artificial intelligence

Today we are increasingly talking about applied artificial intelligence both in reference to very complicated and rare engineering problems born in laboratories “lost” in some high-profile university campus as well as to real situations much closer to every-day life, comprised of simple things for everyone.

There are many examples of AI and we have chosen some of them to explain how this technology can act as an impetus to improve everyday life almost without the user realizing that he or she is being helped, or supported, by an “intelligent algorithm”.

  1. Intelligent Virtual Assistants: at Google I/O 2018, DUPLEX was presented. DUPLEX is a system able to support a telephone conversation with a human being. It can elaborate complex answers in real time and with natural language, without the “human” speaker being aware of interacting with a “digital” speaker. Obviously, to be ethically correct, Google team promptly stated that this system will not “deceive” human speakers and will identify itself as an AI software from the beginning of the conversation. The user who decides to use this system can use it to book restaurants, medical visits, holidays and so on, saving time and automating a set of activities that can be structured within a more or less complex workflow.
  2. More personalized and controlled movements: in 2016 Uber began to use AI (Artificial Intelligence) intensively in many aspects, including accurately predicting users’ travel habits, improving its own maps and creating algorithms for its self-driving vehicles. Furthermore, it focused on systems able to “autonomously” verify the identity of Uber drivers with the aim of strongly improving a very delicate aspect of passengers’ safety.
  3. Crime prevention: nowadays our cities, roads, airports and stations are adorned with cameras able to film any situation. Modern AI systems are able to assist human operators in the difficult activity of analyzing in real-time the enormous amount of videos. Thanks to artificial intelligence, we can predict and prevent fraudulent behavior as well as abnormal behavioral patterns. At present, these systems are able to detect if objects are abandoned, if a person is carrying a weapon, or if someone loiters suspiciously near sensitive places (such as schools, airports, stations and the like).
  4. More and more personalized purchases: the big players in e-commerce like Amazon are experimenting with AI-based systems to try to better predict their consumers’ needs. Their goal is to understand what a person needs before the customer realizes it, perhaps by offering a discount which is impossible to resist. To do this, Amazon analyzes a set of data which includes, for example, the user’s profile and previous purchases.
  5. Artificial moderators and journalists: many online news sites are starting to use artificial intelligence systems both to “moderate” and filter comments or even to write brief articles. For example, The New York Times is experimenting with a system of Jigsaw, Alphabet (Google) incubator, called “Moderator“, which while exploiting machine learning, manages to organize (and in some cases to moderate directly) the comments facilitating moderation by the human operator. Yahoo! and Fox use AI systems too to write some news about  sports, the economy and the stock market.

 

Artificial Intelligence is becoming a part of  our lives more and more every day, as we have seen from these examples of AI which are only a small part of all the applications that artificial intelligence can have. Let’s get used to living with robots because we will meet them more and more often in our daily lives.

Elaborated by Marco Pirrone and Lucia D’Adamo