• Artificial intelligence capabilities tend to be oversold by industry players.

  • The human at the risk of artificial intelligence

    returns to the limits of this technology and addresses its perverse effects, in particular the risk of creating a paranoid society, in the clinical sense of the term.

  • The digital world, made up of surveillance and prediction algorithms, tends to create paranoia among Internet users.

What if artificial intelligence was also causing a psychiatric wave? We tend to praise the merits of AI, to imagine a promising future made up of killer robots, superintelligence and augmented humans. Behind the humanoid fantasy of

Westworld

, there are more immediate questions that deserve our attention. This is the difficult work that

The Human at the Risk of Artificial Intelligence

proposes to do,

published this Thursday by the Presses du Châtelet, co-written by Pierre Rabhi and Juliette Duquesne. The book addresses the limits of artificial intelligence - targeted advertising, algorithmic bias, surveillance ... - and questions the society we are building.

First of all, the term "artificial intelligence" to speak of a machine is overused.

To give an idea, "the most successful artificial intelligences have less common sense than rats", underlined Yann Le Cun, director of research in artificial intelligence at Facebook (FAIR) at Station F in 2018. Talk about computing capacity would be closer to the reality of this technology.

However, the collective imagination continues to attribute phantasmic capacities to it.

The rumor of bugged phones

Behind AI, there are mostly algorithms that stuff themselves with data to improve themselves. And, often, the collection of this information is done on the backs of citizens. Companies spy on internet user behavior and collect all possible data about them for targeted advertising. "In the West, the first surveillance is that of multinationals via targeted advertising", points out Juliette Duquesne. The impression of permanent surveillance creates an atmosphere of paranoid.

And the various scandals in recent years (Cambridge Analytica, Edward Snowden's revelations) tend to fuel public mistrust. It is increasingly difficult to distinguish the true from the false in the practices of the major digital players. Juliette Duquesne returns in particular to an insistent rumor in recent years that our phones record our discussions for targeted advertising.

“I asked the question to 80 people for this survey and, even if it is prohibited, it was not so easy to have a definitive answer, says the journalist who however managed to get an idea.

According to several researchers and associations, the phones cannot be tapped for purely technical reasons: the battery would not hold up.

"The fact of not being able to deny its existence with certainty is indicative of the opacity of the practices", analyzes the journalist.

This also tends to increase the level of paranoia on the Internet.

The acceleration of conspiratorial delusions by social networks

And even if the phones are not permanently tapped for purely technical (and legal?) Reasons, connected objects, such as speakers, or certain mailboxes, are monitored. The book reveals other little-known and extremely problematic practices. According to a study published in 2018, “Some unscrupulous companies take screenshots of their app usage without permission. These captures can contain, not deliberately, personal information such as SMS and be transmitted to the many actors of the advertisement ”. Who wouldn't fall into a form of paranoia in such a context? And it's not over. Social networks themselves play an important role in the acceleration of conspiratorial delusions, through their recommendation algorithms,breaking the last lock of sanity.

Whether it's Facebook or YouTube, these platforms showcase the most “engaging” content, which creates reactions. The more a publication makes people react, the more it is shared and the more it is highlighted. And most people react to fear and outrage, analyzed in 2019 Roger McNamee, early Facebook investor and now repentant. “Fake news is about twice as likely to be relayed on Twitter and six times faster,” confirms Juliette Duquesne, citing a 2018 study. With the idea that anything posted on the Web can being fake - deep fakes are an illustration of this - the virtual world encourages a paranoid vision of society, in the clinical sense of the term.

“Conspiracy is a resignation of thought.

It's going to look for someone else who would explain everything.

We are absolutely certain of everything, there is no longer any need to think, you abandon yourself to the other ”, analyzes Marie-Jean Sauret, psychoanalyst and researcher at the Jean-Jaurès University in Toulouse.

"To doubt everything or to believe everything, these are equally convenient solutions, which both dispense us from thinking", confirmed (a century in advance) the philosopher of science Henri Poincaré.

The digital world pushes humans towards paranoia

Marie-Jean Sauret attributes this shift towards a paranoid vision of the world to capitalism.

Our society "promotes the reign of a scientist ideology which answers all scientific and non-scientific questions by means of science".

Science leaves no room for nuance, it is true or it is false.

“The knowledge of science is essentially paranoid,” he continues.

2 + 2 = 4, that's absolutely certain ”.

However, metaphysics is incapable of competing with the knowledge of science.

The question of the meaning of life does not find an answer capable of resolving all the questions that bother humans during their existence.

Find the Future (s) section here

Conspiracy brings definitive answers where humans cannot find them. It is no longer a question of critical thinking, quite the contrary. "It's a phenomenon of credulity: we want to believe", explained, a few months ago, to

20 Minutes,

Rudy Reichstadt, founder of the Observatory of conspiracy. The digital world is pushing humans towards paranoia in the clinical sense of the term. And this perverse effect of technology is all the more worrying given that, according to 

The Human at Risk of Artificial Intelligence,

AI is not as efficient as we would like to believe. "AI programs, which are very opaque, are often poorly evaluated and continue to be used, even if they appear ineffective or even harmful," the book emphasizes.

“The question we can ask ourselves, concludes Juliette Duquesne, is: is it worth it to create a society that is completely paranoid for AI programs that do not work so well?

There is still time to think about it.

Culture

Artificial intelligence: "In the future, we could see a machine cry"

Culture

Conspiracy: Are we doomed to live in an increasingly conspiratorial world?

  • Culture

  • Social networks

  • Conspiracy theory

  • Artificial intelligence

  • Edward snowden