Artificial Intelligence isn’t dangerous: it’s dumb.

May I speak to Mr. Musk’s parents, please? — said no journalist ever. You might know the guy as a tech guru, an energy clairvoyant or an internet sav source; but asking what your favorite kind of cheese is has very little to do with putting satellites up the stratosphere in order to provide unprecedented internet coverage and speed for the entire freaking planet. I have no need to quote, look at his Twitter (and I sure hope you don’t think I mean it literally). The man can sue me, and I’ll be happy when Justice comes after me and I sit in prison for not paying his lawyer’s established fine in due time, but smiling inside while the internet discusses how he pulled another con-artist move, this time in detail and with a list of victims. In fact, he was accused of sexually harrassing flight attendants and his own daughter changed her name in order not to be associated with him; but nobody talks about that anymore. The spectacle is bigger. Crypto, energy solutions, speed for more con-artists: if we still need oil and gas, we still need geopolitics specialists to say what the economic activity in certain parts of the world has been, and that’s why he and so many other scammers thrive with an obvious lie. In certain parts of the world, nobody’s interested in the truth: they use that as a concept to fabricate the story they want, almost as a sort of pleasure — that thing they were never able to have, maybe because they’ve never worked hard enough to get it, or maybe because people didn’t want them to have it; and what have they been able to give, one wonders? The very thought of the richest man in the world “pleasing himself” makes me want to throw rockets into the first Tesla car I can encounter, but that will be a minor deficit, not a misdemeanor. Should we throw rockets at Tesla cars now? Maybe think about what pleasures the man has been able to give. That’s right: trace Tesla owners, using robots. Woudn’t it be interesting to know who the hell they are, or is that going too far? Maybe we have to ask ourselves if the man alone in this, and we know that is far more difficult to answer.

When I first learned about artificial intelligence, I was studying traditional linguistics in more detail. After my academic advisor invited me to teach a class for the University of São Paulo’s Discourse subject at the School of Humanities (excuse the translation), I was more interested in Semiotics than conversation theory. But that was a mistake. I came to the conclusion that Semiotics in practice was pure marketing, then design, and with design, a lot of programming. That shifted my focus. Conversation theory has big names in Brazil (like Dino Preti, who arguably laid the groundwork for speech recognition, albeit far from intentionally), but so does Semiotics (with Lúcia Santaella’s catalogue, which I’ve never read in full because it was from a different educational institution, but libraries exist for a reason). The problem was the intersections: in a world where conversation goes to Skype, how do we make sense of interaction policy? Facebook’s response, at the peak of demand for robust policy on interactional permissibility, was to invest in Oculus Rift, putting the abbreviation VR all over the news. AR, VR, who cares? We still can’t send a freaking baguette pic (but nothing against the peach). Even WordPress hasn’t figured that out yet, but something makes me think that the company that agreed to follow through with the decision taken by Verizon Tumblr to ban adult content dealth with demand that came from a different source, maybe the part that rules the other 70% of the web. Interesting fact: Tumblr isn’t even mentioned on Verizon’s Wikipedia page, despite it having bought Yahoo’s operating business in 2017. Of course, you’d just have to Google the number of users on Facebook, but I think you can still find that Zuckerberg only earns 1 dollar a year, while his company makes 83 billion during a worldwide sanitary crisis. Another clairvoyant, or should I say surfer? Of course, his intentions are putting people first and profits later, right? Well, of course not. We’ve all watched clips, at least, from the FTC hearings. What the hell are we afraid of?

Artificial intelligence will tell us that they have the perfect model for our business, once we spend time on Instagram talking about it. They just miss the calculation where we realize we’re using the platform when we talk about the business, because the business is freaking ours to begin with. What do the robots say? “show them ads” — but with a logic of “improving user experience”. Well, here’s the big news: conversation theory research would improve everyone’s user experience; but in the wrong hands, it could land innocent people in prison. Facebook, now Meta, came to a point where it’s prohibited to talk about abortion within the platform, and the failure — beyond the ethical, but this time cognitive — is too evident to ignore: they’re associating abortion to violent speech. This data compilation model could only offer its developers (and investors) with compilation, because categorizing would take up too much time (and even the “react” system couldn’t solve the issues to be tackled). It sounds like a joke, but that’s very likely how it went:

1) People are talking about abortion;

2) One of the main arguments about abortion seems to be “killing babies”;

3) Thou shalt not kill;

4) Label abortion-seekers as terrorists.

Without mentioning who funds what, you can always read about the outcomes of these decisions, and the intricacies might puzzle some, as I’m sure is true for reporters at The Intercept covering the case; but the stories are coming out. Of course, it’s an ongoing debate; but The Verge has covered the Biden administration’s response to the Meta’s policy and other should follow. What’s harder to tell is how far articifical intelligence will go in making assumptions about people. It seems almost as if it’s designed to look for evil, not for good. A mechanism to dig dirt on opponents. Sound familiar? Maybe it’s because it is. We all remember Hillary Clinton won the popular vote in the United States in 2015, and the campaigns against Biden in 2020; we also remember how Lula was put into jail by prosecutor Deltan Dallagnol, who investigated Petrobras, a company that was going to put a major part of its profits to invest in Education and Health in Brazil. The problem is that few people connect the dots for a living — or in order to live. And while that might be a young writer’s exaggeration, we are absolutely not talking about competition here, unless it’s the nature of democracy to compete with authoritarianism — on different levels, but globally.

Image: Pexels

Leave a Reply