An Ameca model humanoid robot from the British company Engineered Arts and equipped with artificial intelligence is displayed at CES 2022, on January 5, 2022, in Las Vegas, Nevada
©Ethan Miller/GETTY IMAGES NORTH AMERICA/Getty Images via AFP
Google, Meta and OpenAI are investing heavily in this technology, which increasingly captures the public’s imagination. More and more researchers believe that now is the time to reconsider our expectations of artificial intelligence.
Atlantico: How much do we overestimate the power and capabilities of artificial intelligence? And in particular his ability to become conscious?
Fabrice Epelboin: This is actually based on the difference between digital and digital. Digital (programs, microprocessors, etc.) is a world in which bathe what we call, on the side of the media and politicians, experts, those who know how to code or who understand how it works and who have an intimacy with the technical thing. The digital, on the other hand, relies on the mental representation that people who are not intimacy with technology have in order to apprehend things that could be magic or myth.
For most people and for almost all journalists and political decision-makers, this world of technology and artificial intelligence is only accessible through mythology, such as beliefs in deities like Zeus, who sends lightning, or Athena, who starts wars.
Through an oversimplification, through metaphors that make it possible to apprehend concepts that are too complicated, society vis-à-vis technologies and artificial intelligence is today split in two with on one side people who are in digital and those in digital. Some people are able to make the connection between the two.
Most people’s apprehension of technology is not necessarily very rational.
All these technologies are built with words, which are the responsibility of communication between humans. The words that will be used (networks, intelligence, data, etc.) will allow people who are intimate with technology to know what they are talking about because they need conventions to designate things. It also allows you to create bridges with people who do not have this knowledge and to whom you will have to tell the myths to generally sell them something, either to write an article to generate attention, or to sell a project. law or to sell technology to a large corporation. For this, “storytellers” will intervene to pitch the technologies. They don’t necessarily have a crazy intimacy with technology but they know how to tell stories.
Artificial intelligence: Google puts an engineer on forced leave for claiming that the group’s chatbot had its own intelligence. But why ?
And in this context comes the object of fantasy, artificial intelligence, the ability of man to build an intelligent being. This is part of the most tenacious fantasies and myths such as Frankenstein. This is a founding myth of the human being who suddenly plays at being God. Inevitably, vis-à-vis people who only have digital as background and who do not understand digital, it is easy to go very far in phantasmagoria. We face this myth with artificial intelligence.
AI is something very vast. Marvin Minsky was one of the founding fathers of AI and an excellent popularizer with his books. Artificial intelligence is an old topic of computer science research. There are a multitude of branches. Today, some people are therefore doing code, deep learning. Over the past fifteen years, one of its branches has given rise to an industrial application that will affect people in tune with digital technology. Previously, artificial intelligence only concerned digital people. Individuals who did not understand the technical thing had vaguely heard of this technology but it had no concretization. There was no myth, no story to tell. It only interested enthusiasts.
For the past fifteen years, we have witnessed the development of one of the fields of artificial intelligence, deep learning. It has become an everyday object. This technology recommends a Netflix series that you might like. It allows Amazon to recommend a product to you and is present in the advertising engines whose challenge will be to anticipate your desires, your needs and your requests. This is just statistics. There is no form of emotion and intelligence in the mythical and deistic sense of the term. There is nothing equivalent to what one might call life. But you need myths because it costs a fortune.
Why We Might Have Digital Twins In The Years To Come
There are very important technological, industrial, financial and societal issues behind all this. Recourse to myth is therefore privileged. So there’s a fundamental schism in society between people who know and people who think they know and actually live in a mythology.
Artificial Intelligence is the strongest iteration. We really touch on the myth of the man who gets closer to God by creating something alive himself.
Is the money that is invested in these technologies justified with regard to their current capabilities?
One of the most glaring promises of artificial intelligence is predictive, abilities to predict the future. This may concern your future choices of films and series on streaming platforms. Another area concerns health. With all the health data that you have entrusted to the State, and that the State has transmitted to Microsoft, it will be possible to compare them with millions of French people. Through the history of these people, it will be possible to anticipate the evolution of your state of health. This may have negative consequences in some cases with the fact of no longer being insured if you risk, for example, dying of cancer.
Artificial intelligence requires considerable investment. Typically in the field of insurance, we are going to witness a radical change. This could swing almost the entire market towards players like Microsoft and Google. They could carve out the lion’s share of tomorrow’s insurance business. They will decimate the traditional players who are unable to accumulate so much data and who do not have this technical expertise. The insurance market is huge. This justifies that colossal sums are invested in artificial intelligence. But this is just one example of markets that could be totally disrupted by this ability to predict the future using artificial intelligence.
Is it still possible to escape this digital hell in which we let ourselves be locked up?
These technologies bring and promise colossal profits. The evolutions are even difficult to apprehend in the face of the upheavals generated by artificial intelligence.
For health insurance, we risk witnessing a transition from a rather egalitarian system of solidarity in the face of illness to a new form of inequality. Whatever your income, if you are affected by metastasized cancer, you may be rich, but you will no longer be insured because your treatment costs a fortune and you are not “profitable”. We do not yet know this form of inequality. It will be introduced by technology. This will be very interesting to observe because it will strike individuals who until now believed themselves above inequalities. So we’re going to move on to a biological inequality with artificial intelligence.
Artificial intelligence is an extremely promising sector that justifies the investment of completely crazy sums. But it is essential not to fall into myths and legends.
What is the part of the lack of lucidity on the AI? What about cynicism and marketing tactics?
If you look at the National Assembly, less than 1% of MPs truly understand technology. The Assembly will however decide on fundamental laws on cryptocurrencies or artificial intelligence, which are areas that will have to be regulated. They are going to do it according to their vision based on the mythical aspect of this technology. This situation is catastrophic.
Lobbyists will also tend to highlight the myths and legends aspect of this technology. The references are numerous in literature, in science fiction, in the original myths in relation to this technology. The story telling is endless on topics like artificial intelligence. It is then very easy to influence and manipulate anyone.
Should we “reset” our expectations of artificial intelligence? Demythologize our expectations on the subject?
You just have to be modest. We live more and more in societies in which individuals are less and less educated and have less good knowledge. 30% of Americans are convinced that the Earth is 7,000 years old. 75% of French people are convinced that nuclear energy emits carbon dioxide. This testifies to the stalling of intelligence and knowledge. There was a dropout. We have lived for a long time, since the Age of Enlightenment, on the idea of constant scientific progress and constant progress in the scientific understanding of the population. Very clearly, this is over, including at the highest level of the State. These people have in their hands the management of very large companies or gigantic political entities and they do not understand anything about this technology. They nevertheless perceived that this represented the alpha and the omega of the future of the planet, but they do not understand anything about it. This does not prevent them from making decisions in the matter. These people are the main targets of storytellers. It is therefore essential to learn and understand the reality of this technology and how it works.