By requesting the removal of copyright to ensure respect for the “right of machines to learn”, the president of Microsoft decides, according to Vincent Lorphelin, to promote AI rather than human intelligence.
Brad Smith, president of Microsoft, is committing a moral wrong against the society. Talking about generative AI like ChatGPT, he says, “ We all have the right under copyright law to read and learn. We now wonder if we can enable machines to learn in the same way. I think there is a societal imperative to make this possible. “
This intervention may seem trivial. After all, if machines absorb the internet to create articles and illustrations at a ridiculous price, why deprive yourself of it? Are the New York Times and other rights holders who oppose this use of their content not making a bad case against Microsoft? If machines don't plagiarize, what's the harm? This is the unacceptable purpose of this speech.
An advantageous legal vagueness
We know that GAFAM platforms harvest a mass of small contributions and promote free work. They change again in dimension with AI, whose voracity requires the conquest of new territories, those of the authors.
As an exception, the legislator had defined “fair use” to authorize Google to publish extracts of private content. With the legal outcome still uncertain, former attorney Brad Smith knows public opinion will play a big role. This is why he distills the adulterated mental image of “the right of machines to learn”. Its “societal imperative” is only a vulgar contestation of the legitimacy of the authors to defend their territories.
He diverts attention by debating falsely ethical issues. Is human intelligence doomed to be downgraded? Can we let AI make decisions? Although he may adopt the posture of the wise man in the face of a less civilized Elon Musk, they are both augurs of the same illusion: that of surpassing human intelligence.
AI: an illusion of superiority?
The best novel will never replace life experience. The best videoconference will never replace a real meeting. However perfect it may be, the model is not reality. Each of us has an intelligence of the world that will always exceed that of machines, which only have intelligence that of its models.
AI is enlightening our intelligence. They allow the environmental police to detect polluters, the teacher to understand the difficulties of a thousand students, or the laboratory to anticipate the effects of a drug.
Spreading the illusion of the imminent superiority of machines is irresponsible. It is to deny strictly human work, which draws its inspiration precisely from the world and its models, when the artist reveals something unsaid, when the driver compensates for poorly responding levers, when the doctor doubts a diagnosis that is too obvious, or when the judge forms an intimate conviction.
Defense of the authors
Feeding the illusion of a competition between intelligences and attributing human rights to machines is an offense against society. On the contrary, we must affirm the pre-eminence of human intelligence and guarantee the integrity of its territories.
Brad Smith's illusion would not have any gravity if it were not the work of the world leader in AI. Won by this illusion, the Interministerial Commission on AI, which has just submitted its report to Emmanuel Macron, affirms that machines will surpass humans. It defends, fortunately, the fair remuneration of human creation, but without realizing that it is the very appreciation of this fairness which is manipulated.
Europe has already decided with the AI ACT to protect copyright. It must now establish the land register of the territories exploited by AIs and defend their rent. It is up to France, starting with the mission just launched by Rachida Dati, Minister of Culture, to find the equation that will develop AI and copyright together. Our collective is also committed to using a patent essential to generative AI, of which we are the rights holders, to put our grain of sand in Brad Smith's infernal machine.