People Mauro Martino


IBM Research AI, Northeastern University

Mauro Martino is founder and director of the Visual Artificial Intelligence Lab at IBM Research based in Cambridge, U.S.A., and Professor of Practice at North-eastern University in Boston. He is a pioneer in the use of the artificial neural network in sculpture. His interactive AI works are used by millions of people, some examples are: Watson TED (, News Explorer (, Forma Fluens (, AI Portraits (

His works were published on important scientific journals and quoted in several conferences such as Nature, Science, PNAS, NeurIPS and on various books on data visualization: "Data Visualization", "The Truthful Art", "The Best American Infographics" published between 2015 and 2016.

Mauro is a multiaward-winning designer whose projects won the gold medal at The 2017 Vizzies Visualization Challenge of National Science Foundation, the Innovation Award by Design Award of Fast Company and the Kantar Information is Beautiful Award. His projects were exhibited in international festivals and exhibitions, among which Ars Electronica, RIXC Art Science Festival, Global Exchange at Lincoln Center, The Serpentine Gallery in London, 123 DATA, Paris.


How many times do we repeat ritual phrases to other people? How many times do we try to communicate our feelings in depth and really understand others? Empathy and synthetic humanity will keep us awake from the emotional numbness into which we very often fall.

— Mauro Martino
The interview

A conversation with Mauro Martino

People talk a lot about artificial intelligence and many are afraid of it, but how can we define this revolution in a few words?

As a field of scientific research, artificial intelligence is gaining a lot of attention. I think that very soon we will see this discipline growing in all university faculties, from law to medicine. Of course, there is no artificial intelligence equivalent to human intelligence. We often misuse this term, but in reality it’s basically statistical models that are able to process unstructured data in real time. The shocking thing is that we do not need artificial intelligence to replace most of the work we do. I think that fear comes from here. Dante wrote “fatti non foste a viver come bruti, ma per seguir virtute e canoscenza” ("you weren't made to live like animals, but to follow virtues and knowledge"). We have forgotten to raise our spirits, we are living like animals, doing repetitive and stupid jobs.

The modern theory claims that the only limit of AI is the lack of empathy and humanity, but is it really a limit?

Empathy and humanity can be simulated. The reading of body movements, micro facial expressions and voice can be interpreted with accuracy. We can train mathematical models to recognize feelings related to these data and to react with actions, and words that change the current emotional situation.

How many times do we repeat ritual phrases to other people? How many times do we try to communicate our feelings in depth and really understand others? Empathy and synthetic humanity will keep us awake from the emotional numbness into which we very often fall. We are building machines that simulate and are not, that pretend and do not know. Our pride will drive us to be less ordinary and predictable. Always creative! Always human!


What is AI's present? How could a designer use AI today to increase his productivity and creativity?

Little by little we are moving towards artificial intelligence, we are at the beginning and no one at this stage can be optimistic or pessimistic. So far we use models that learn from the real world, watching and listening. They learn to recognize and better reproduce objects, situations, emotions, words, sentences. Now it's difficult and expensive to educate them, and different models don't interact well with each other.

The best results are probably in the field of vision, that is why designers should be the first to use this technology. Thanks to GAN (Generative Adversarial Network) models, we can play with shapes and colours, merging them and blurring the boundaries between different classes of objects. AI is the new muse of the designer.

In the project we need visual cultural references to conceive an evolution of the created object. Thanks to AI I can create artificial images of all that exists and does not exist, I can be guided in daring and unusual visual paths. The designer has to experience this multidimensional visual universe. The latent space, in which the images generated by the GANs move, seems to have been created for designers, as an inspiration for their creativity. It's not just about passive experiences, I can control images generated by AI through my design gesture. Drawing, the primordial form of design expression, becomes the magic stick to direct the overflowing production of AI images. I think there can be a perfect dialogue between those who control the design sign, seen as the ability to design through drawing, and an AI that stores in itself every possible shape. Example: the designer starts designing a new dress, a shoe, a table, a chair, a poster, a website, a storyboard for a film... AI observes the drawing and gives suggestions on how the object could be; it recreates it as a hyper-realistic rendering, and shows different versions and variations. The more accurate and detailed my drawing becomes, the more hyper-realistic suggestions and previews follow my style and aesthetic sense. The system can be trained on my collection of drawings, renderings, photos and capture my design style and start suggesting shapes and colors that are increasingly consistent, or dissonant if I want to understand what I will never draw.


© — All rights reserved. — Published on 20 March 2019