Focus

From machines that learn, a new industrial revolution will be born

01 May 2023, by Michele Chicco, translated by Alex Foti

Generative AI and sensing technology: they are the two elements that will radically change society, explains Nader Sabbaghian, General Partner of Capital 360 and Bocconi Fellow. But artificial intelligence already enables a lot, as the 6 startups incubated by B4i tell us and as Mazet Sonilhac points out when looking at venture capitalists

Nader Sabbaghian The computer science we have known until today will no longer exist. The genie is out of the lamp: "Artificial Intelligence (AI) is a new technological paradigm that will take us to the next industrial revolution, a bit like what happened in the early 2000s with the generalized diffusion of the Internet". And Nader Sabbaghian was there: in 2001 he was part of the founding team of Bravo Solution, today he is General Partner of Capital 360 and Academic Fellow at Bocconi University, where he teaches Technology and Innovation Management, and is faculty member of the MBA of SDA School of Management. "Classic computer science – he says – is deterministic, since it allows you to automate processes only by way of algorithms. AI is something altogether: machines acquire cognitive abilities, learn from mistakes and improve over time". VC funds are heavily betting on this: according to Dealroom data, $167 billion went to AI startups between 2021 and 2022. In Italy, it's $401 million in 59 rounds of investments.

What are the prospects for the market?
Huge. There is the opportunity to create new business models and new startups, but with artificial intelligence we can also improve business processes. AI works only if we have high calculating speed and vast troves of data, essential conditions for solving enormously complex equations at affordable costs. AI is now entering everyday life, and at the speed the market is going within the next decade we will see mass products and services generated by artificial intelligence.

On what innovators should focus to attract capital?
They must identify concrete solutions for very specific problems at hand. Much of the entrepreneurial challenge is to identify the specific segments that can rely on big data to train the models. We are keener on Narrow AI, i.e. startups that are well focused on their target: two years ago we invested in Pallon, a Swiss company that automates the process of inspection of sewers. A complex activity, now made with probes that track the ducts and are then viewed by a human being. A slow activity that can be done much more precisely with AI, once it learns who to recognize leaks in the sewage network. Pallon exploited the first customers to collect data, but now after thousands of kilometers inspected, AI has learned enough to start replacing humans and do this work independently.

What is the teaching for the founders?
Better to be the best in the world to do a super-specific thing than trying to achieve a great ambition. We know how difficult it is: the founders that come to us with an idea often do not yet have even the product prototype and if they aim is too wide, they will struggle to achieve what they want.

What should aspiring entrepreneurs focus on?
There are two really interesting themes that will allow you to radically change society: the development of advanced sensors and generative AI, the creative artificial intelligence that originates from ChatGPT.

Sensors, therefore hardware. What is the connection with AI models?
Problems need to be solved in the physical world, not only on our computers: we will use artificial intelligence to move from a world of automation to a world of autonomy. Today we have very sophisticated robots that have no cognitive skills; to have AI rise in their stead we need a major development of sensing technology. You have to invest a lot in sensors suitable for creating machinery that can navigate this new world like we humans do.

Generative AI enables a computer to reason like us.
I didn't think that I would be able to see something like that in my own lifetime. We are building the conditions for machines not only learn what we know to do, but go further that hat. Until now, AI didn't do it, it was merely a problem solver. Generative AI emulates the act of human innovation, since it combines previous knowledge to create something new. We are still at the beta stage, yet software like ChatGPT is already giving the goosebumps. What matters is the basic technique: these programs will soon be used for scientific research because they will be able to take in all knowledge, contextualize information and transform it into something new.

The AI will revolutionize entire economic sectors: should we be more frightened or fascinated?
The next twenty years will be very interesting from the point of view of economic development, there will be another industrial revolution. And like all past ones it will cause major upheavals. But you cannot put the brakes on technology: now it has emerged and we must understand how to adapt our way of developing human resources to the rules of the new world. Human beings know how to adapt well, and we will do so now too, and learn how to do business in the new conditions. Governments will have to make sure that this technology develops in a non-destructive way, and will be called upon to protect those who, during the transition, are likely to suffer the most from the disruption caused by AI growth.