“Why robots won't replace humans,” explained by Inbenta Chief Jordi Torras

Artificial Intelligence (AI) is progressing at an insane pace. Every sector appears to be going ‘smart’ by integrating AI into their products. Robots, drones, music players, smart cities, medicine, you name it, AI is finding a way to creep in. This sudden advancement has led some to fear the day when AI will take over human jobs and ultimately, end up replacing humans at the top of the food chain.

But Jordi Torras, CEO of Artificial Intelligence firm ‘Inbenta’, believes otherwise.

Where big names like Elon Musk have warned against the use of AI, Torras on the other hand regards these warnings as little more than ‘myths’. In an exclusive interview with Business Recorder, Torras presents his alternative view.

Q: What makes you think AI won’t take over human jobs?

JT: Artificial Intelligence and its many use cases do not go without risks, but some of those risks concern things that are happening today -- whatever AI could bring to humankind, it’s already happening now.

It’s all part of the greater Industrial Revolution. Humans had been trapped in boring jobs, and when you look at it now, we have better automation, faster computers, and smarter algorithms. In particular, AI is going to take over these boring jobs that in the past had to be performed by humans. It’s going to give space for us to have a more creative role in the world and to focus on jobs that better exploit what computers are bad at, which are creativity and empathy.

It is not about bad or good AI; it’s the use that we do with this technology. For example, atomic energy can be used in very creative ways or in devastative ways -- it’s all about how we use it. I am not afraid of an apocalypse... robots killing everyone? It sounds very far off.

Q: Do you think society is ready for Artificial Intelligence?

JT: We are now entering an era that is going to be as transformative for humankind as the Industrial Revolution. Similar to how humans began to use larger animals for farming, or how they progressively built more powerful machines to increase workforce productivity, AI is bound to make similar strides, but confined by our own mental capacity. Eventually, we will create machines that will perform things that we don’t want to spend time or effort doing ourselves. This is how AI will spark the next Industrial Revolution -- the actual name for that is yet to be coined, but this revolution will bring extensive change that impacts what we do in our daily lives.

Q: How can people prepare themselves for AI and the machine learning revolution that’s coming?

JT: We live in a world where you start as a kid, you go to school, you go to college, you learn some skills, and then you are supposed to get a job that will buy you a house and a car, be able to afford a family, have kids of your own, and so on. That’s the mental and societal framework we all share. AI will change everything, starting from education to the way we look at our professional lives, our expectations of ourselves, and how we manage expectations of our income. All of that is going to change, but it won’t be immediate. We have time to prepare and to understand better that, if your life goals include pursuing a job that is “boring,” think twice, because that job might not exist by the time you reach it. New and more interesting jobs will arise, and the cost of completing tasks using robot or AI technologies will be virtually eradicated, or the cost will shift somewhere else, and no one will spend more on automation the same amount they would spend on human jobs.

With the rise of AI, we should all work to ensure that it is first used in the right way. It’s up to us to determine that, and no one else is going to make that decision for us.

Q: What will be the impact of AI and the machine learning revolution in developing markets, many of which still lack basic infrastructure?

JT: Look at Africa, for example. We have this image of Africa as a poor place, constantly developing, and having an unacceptably high child mortality rate, but if you look at actual data from the last 15 years, the state of Africa has improved. Mortality rates are not as severe as they used to be. Of course the internet is not as strong or widespread as it should be, but cell phones are overwhelmingly prevalent. Everything is relative, but what’s happening in Africa is that people bypassed the internet and jumped directly into mobile, creating the first era of modern technology for them. Although truly bridging the technology divide will take a while, with the right use of technology and the right policies implemented in developing countries, we’ll start to see greater improvements. The only things that might stifle development in these countries will be politics, policies & regulations, institutions, wars, and other crises, but differentially technology will be the mechanism that will help these places to prosper and grow exponentially.

Q: In a recent documentary tech entrepreneur Elon Musk warned that “god-like super intelligence can take over the world.” What would you say to that?

JT: I conceptualize AI as a pyramid with five stages. The first and most basic stage is simple computing. Show a modern calculator to someone in the seventeenth century and they will say, “Wow! This is AI!” However, that machine is only as “intelligent” as us because we spent 10 years learning multiplication and division -- this machine is just taking what we learned and computing it. The next level is automation, which is happening more frequently and has been replacing and transforming within society for many years. The third level is specific AI, in which a machine is now taking care of what humans previously had to handle themselves. The fourth level is general intelligence, the level at which AI becomes as intelligent as a human in everything we can be intelligent in. The last stage is super-intelligence, intelligence that is above humans.

Today, we have specific intelligence, we don’t have general or super-intelligence. We’re not there now, and I believe we’ll never truly have it. never going to really have it. Ten years, maybe one hundred years from now, perhaps, but groaning about something that may never happen is useless at this point.

Q: What if we achieve general or super intelligence? Would it be difficult to contain and could it eventually become a threat to us?

JT: If we have general intelligence, no one would be able to control it, just as you cannot have a human under control forever. If it is super-intelligence, which is not going to happen, then we can’t imagine the benefits or consequences. It could be dangerous, or it could be what we want it to be and more.

It’s too far out to determine. I say we shouldn’t over-investigate the “dangers” of AI -- and regardless of whether Mr. Musk can control it or not, it’s going to happen anyway. So, instead of trying to stop it, let’s focus instead on the positives and negatives.

We can’t stop or reverse progress. Mr. Musk is working on driverless cars, so that is a risk, but I don’t think the alternatives are here. People die every day on highways, anyway.

Q: How would you describe the future of AI over the next 2-3 years?

JT: I believe that AI has been here for a long time, and we are going to see great progress in many areas, such as natural language processing, computer visions, self-driving cars; those will all evolve. But at this point, there is a little bit of hype and over-coverage of the whole thing. AI is not the hottest thing anymore, blockchain is. And everyone says, “Oh! Blockchain is going to change the world,” and a few months from now, people will forget about AI. All of this paranoia about AI taking over the world is going to sound old-fashioned. The technology is still going to be there, but my prediction is that the next crazy thing is going to be blockchain, not AI.

It’s always about what’s in fashion. First, it was semantics, then came digital intelligence, then there was the Internet of Things (IoT). People went crazy over all of those, but they aren’t cool anymore. Despite progress to be had in the coming years, sooner or later, AI itself will begin to sound very outdated.