Getting a little help … from AI

I don’t know about dogs or other animals but humans have an uncanny ability to recognise "things" in shapes found everywhere. Ever saw a cloud reminding you of a face, of a horse of a… whatever. Yes I am sure you have.
Yet recognising shape is really tough for computers, so tough that image recognition has been a sort of holy grail and a test for computer smartness.
This is (another) area where we have seen amazing progress in the last years. I remember just 15 years ago saying in some of my presentation that it was not possible to ask Google for an image of a "barking dog" and at that time it was true (you could get an image of a barking dog if the image was associated to a caption saying "barking dog" but Google was not able to tell if the image was containing a barking dog).
Now we have available (for free!) a service from Google to help us drawing images on our computer, tablet, phone…You draw a sketch of what you have in mind (see the masterpiece I draw, first figure) and Google recognise your chicken scratch and suggests some interpretations. You click on the one you want and … voilà, there is your nice drawing (see the second figure).
The magic is in the capability to interpret a few lines as a representation of an object. This is done using some sort of artificial intelligence. More specifically, Google is saying that they use machine learning, which is a great trick because the application will keep learning, and getting better, the more it is used.
Google started releasing some games in November 2016 that actually helped in training the application. This is really smart if you think about. You create a learning machine and let it go free to learn. Potentially all of us become teachers, and the more teachers there are the faster and the better the application becomes. It never tires out of learning (it is not like us that very soon would lose drive and focus).
If you try it out I am sure you will love it, but as you get entertained just pause a bit to think at the amazing intelligence that is supporting this apparently simple application. There is nothing so difficult to recognise that I meant to draw a cat, isn’t it. And yet it was simply impossible for a computer to do that just few years ago. This is marvellous and scaring at the same time. We are entering a time where computers can be as good as we are in many practical, everyday things … just way better.

About Roberto Saracco

Roberto Saracco fell in love with technology and its implications long time ago. His background is in math and computer science. Until April 2017 he led the EIT Digital Italian Node and then was head of the Industrial Doctoral School of EIT Digital up to September 2018. Previously, up to December 2011 he was the Director of the Telecom Italia Future Centre in Venice, looking at the interplay of technology evolution, economics and society. At the turn of the century he led a World Bank-Infodev project to stimulate entrepreneurship in Latin America. He is a senior member of IEEE where he leads the New Initiative Committee and co-chairs the Digital Reality Initiative. He is a member of the IEEE in 2050 Ad Hoc Committee. He teaches a Master course on Technology Forecasting and Market impact at the University of Trento. He has published over 100 papers in journals and magazines and 14 books.