The world of tomorrow will be divided -laborally speaking- between those who compete against the machines and those who build such machines. And the former are doomed sooner or later to run out of job. This was warned by the economist and professor at King’s College, Daniel Suskind, which has closed this Wednesday the First Catalan Congress of Treball. At the height of artificial intelligence, commercially led today by the GPT Chat, this Briton who has advised the prime ministers Gordon Brown and David Cameron alert that no one is safe from automation, even the most skilled workers.
How long will it take for this interview to be done by Chat GTP and not a human like me?
I have bad news for you, because this interview could already be done by Chat GPT. Nowadays there are economic forums in which prestigious experts share a panel with the GPT Chat.
So your job is also at risk…
Yes, totally. GPT Chat is fascinating, it opens a new chapter of history. Technology every day, slowly but inexorably, will be more capable of doing jobs that until now we thought could only be done by humans. Interview someone, drive a car, operate surgically… I think it will be one of the defining characteristics of our age.
“The key is how and who measures the quality of artificial intelligence work”
I will rephrase the question. When will GPT Chat do this interview better than a human journalist?
It puts me in a complicated situation, because the journalist is in front of me. [Risas] Currently there are systems that perform highly sophisticated medical diagnoses. How do we judge the perfection of that diagnosis? We have a problem there, because programs also make mistakes. But if we compare it to a human doctor, they can make many minor mistakes. How to measure the quality of artificial intelligence work and who measures it are the key.
You defend that the machines have not come to take away our work, but in many cases to make us more productive. But they have also generated precarious jobs that did not exist before. I’m thinking about the delivery people of the food delivery applications.
The direction that technology takes is a consequence of the decisions that business and political leaders make. If all the precarious employment that has emerged is because the regulation has allowed it. We are in a very interesting moment in which from the public we have to think about how to redirect technology more actively towards better jobs. For example, in the United States, the tax system favors buying a robot more than hiring a worker. And those taxes could be redirected to do the opposite, there is a very interesting debate there.
“The British royal family torpedoed many technological advances for fear of the effect it might have on their domestic service”
In that sense, do you see the potential of a Luddite party [contrarios a las nuevas máquinas]who defends limiting technology to preserve the employment of more people?
It is totally doable. In fact, that was, in part, one of the arguments used against Donald Trump, questioning why he focused more on criticizing that immigrants took jobs from Americans instead of focusing on artificial intelligence that also supplants workers. If Trump were a true populist he would have focused more on the latter.
In Europe can it be a reality in the short term?
Completely. This is what has historically happened with technology. The British royal family, for example, torpedoed many advances, fearing the effect it might have on their domestic service.
Do you see the irruption of this type of party as dangerous?
I see it more as a mistake. I am optimistic about technological progress and think we need more, not less.
What do we do with the people who can’t become the builders of the machines?
With machines we can compete or complement each other. The latter will be the ones to prosper and the former are doomed not to. The challenge comes when there are jobs to fill but for different reasons there are people who cannot do them, one of them is the disconnect between needs and abilities. It is a very serious problem and unless we respond efficiently the social consequences will be very negative and social discontent will grow. More education and training are essential.
“People should be debating the 4-day work week”
Can the 4-day work week be a resource to redistribute tasks in a world with less and less work?
We are becoming more and more prosperous and the question is how we distribute that prosperity. Through a basic income or that people continue working, but fewer hours? It’s a fascinating question to which I don’t think there is a definitive answer. They are dramatic interventions that can only have legitimacy if the people support them. It’s up to each political community to decide, and the four-day work week is exactly the kind of issue people should be speaking out on.
Are they incompatible?
Not necessarily, it depends on how ‘basic’ the basic income is, for example. That is why it is so important that there be a public debate on these issues. Personally, I bet more on a conditional basic income, that is, that in exchange for receiving something from the common fund, people have to provide certain services in exchange. The difficulty comes when if this has to be a solution to the lack of work, how are people going to be able to contribute if they don’t have a job and an income?
“Taxing the use of robots is complicated, how do we count them?”
Does it make sense that robots or ‘apps’ pay taxes?
This idea has several problems, the main one being how we measure how many robots or programs a company uses so that we can later tax them. It’s not as easy as counting employees… The spirit of the idea is fine. At a time when technologies will become increasingly important, pricing the owners of said technology becomes more necessary.
Different studies and sentences have determined that the algorithms are not neutral and reproduce prejudices or stereotypes of their creators, including the administration. How to minimize those risks?
The main risk is that most of these biases are unknown to us. The first thing we need is to be aware of this problem and involve political and business leaders. And demand more transparency regarding those algorithms.
“Administrations are not transparent enough with their algorithms”
Related news
Do you think that the Administration is currently sufficiently transparent?
No, it is one of the main challenges. Currently many of the new technologies are not based on human reasoning, but are learning from experience and that makes them especially difficult to understand.