The physicist and researcher in AI (artificial intelligence) Max Tegmark affirms that we do not need to worry about whether one day intelligent machines will dominate us: “They already do and they have not had to be too much.” And the truth is that he is right. I’m not talking about the current “spring” of AI, or social networks, or even the internet. Machines didn’t have to be very smart to rule the world, or have we forgotten Y2K?
The 2000 effect was the fear that at the turn of the millennium the information systems would stop working due to an error in the storage of dates. It turns out that at the beginning of time the programmers had decided that two digits were enough to save the year in a date – it was the fifties.
Perhaps we computer scientists, who are by nature disbelievers, did not have enough faith that our information systems would last until the new millennium – or perhaps that it was the human species who did not.
It was called “the problem of the millennium”. The most apocalyptic said that if we did nothing, the planes would crash, the banks would calculate the interest since 1900, and what was worse of all, the video games would not start! The world spent more than 308 billion 90s dollars to avoid the apocalypse. Machines ruled the world by being anything but intelligent.
Now they are not, but it seems so to us and that is how we qualify them. Just this week, in addition to reading countless articles on generative AI and even more countless “interviews” with ChatGPT, I have participated in a symposium of the Diputación on culture in which AI has monopolized a good part of the conversation, the popular culture magazine Canemàs has presented a dossier on culture and digitization, Time magazine has a dialogue with ChatGPT on how to design the cover on the cover, and last Friday at CosmoCaixa the pianist Marco Mezquida offered a concert where an AI responded to his improvisation and vice versa.
It was the same Time magazine who, in a recent investigative report, showed us the B-side of fashionable AI. Behind ChatGPT there are also armies of low-paid workers in Kenya who review the conversations and tell the machine what has responded well and what has not so that it learns. This allows us on this side of capitalism to generate poetry in the style of Lord Byron or images in the style of Dalí.
Now, a quarter of a century after the “millennium problem”, it seems that we are back in the fifties when scientists and the press referred to computers as “electronic brains” and 2000 was still the future.
As then, today we talk about artificial intelligence, intelligent machines, their creative capacities and their impact on society. For the moment, as always, the party is paid by Africa. Surely, before 50 years we will all have to scratch our pockets again to fix what we can’t even imagine now.
