Technology has a paradoxical effect: it delivers results that satisfy our desires while diminishing our abilities. Agriculture provided vast amounts of food, causing humans to stop recognizing countless edible plants, GPS ensures everyone arrives on time even at the most difficult places despite being the least navigationally skilled generation in history, and writing reduced the capacity to memorize. Now, a group of 25 Fields Medal winners (the Nobel Prize of mathematics) warns in a letter that artificial intelligence could amplify the corrosion impact on human capabilities through irrefutable results in mathematics, one of the most exalted human activities, emptying them of meaning.
Read more Antonelli celebrates Norris’s jinx in Madrid
The letter is a response to what seems to be the first of several announcements of advances made by AI in solving some of the most complicated problems in mathematics. Last Tuesday, September 8, OpenAI claimed to have solved the Navier-Stokes equations, one of the Millennium Problems that the Clay Mathematics Institute selected in 2000 as the most difficult and important mathematical puzzles.
The letter warns that the way of working that provided these spectacular results confuses the means with the end. “Solving problems is only a tool and a vehicle to reach the main goal of conceptual understanding and deep knowledge,” they write. For mathematicians, solving these famous problems was never the goal itself, but a sign that there were new ideas and methods worth studying, discussing, and transmitting. According to them, “the mass production, at an increasingly rapid pace, of true or false claims could destroy the fertile ground instead of giving life to new ideas.”
Another problem pointed out by the Fields Medal winners is that the chain of knowledge transmission is broken, a concern shared by intellectuals in other fields. Mathematics, like other academic disciplines, advances through the exchange of ideas in talks and discussions and through the training of students who receive the torch of knowledge to continue advancing into the unknown. “Without the willingness of mathematicians who must take charge of its development and integration into the mathematical canon, ideas conceived by AI would never fully come to life and the crucial human transmission chain among mathematicians would be lost,” they warn.
They also criticize the way results are announced within the frantic race for hegemony among companies developing artificial intelligence. “Often, these solutions are announced hastily, leaving no time for proper writing, isolating new methods and ideas, nor citing relevant prior work by others. As in all creative professions, this raises serious dilemmas of attribution and plagiarism,” they point out.
Read more Valencia makes a major change and dismisses Carlos Corberán and CEO Ron Gourlay
The resolution of the Navier-Stokes equations has been a perfect example of these dilemmas. This week, one of the mathematicians at the forefront of solving the problem, Tristan Buckmaster, accused OpenAI of trying to rush the publication of their results and exclude his collaborator Levent Alpöge from authorship because he is linked to their competitor Anthropic. Moreover, in the whirlwind of the tech company’s announcement, it was often overlooked that the mathematical core of the breakthrough was in the work of Spanish researchers Diego Córdoba and Luis Martínez-Zoroa, whose original method was the indispensable foundation for the progress of Buckmaster and Alpöge and the final OpenAI result.
The authors see a much bigger problem in this case. In many intellectual and creative professions, ability, years of training, and effort not only serve to produce a final result but to develop the capacity to understand and formulate new questions. If AI can produce these results directly, the formative purpose is lost, and one can have many answers while having an increasingly poor understanding.
Technology has always posed humanity with the dilemma of means and ends. Results are an essential motivation, but if they appear as if by magic, as the Fields Medal winners remind us, they empty our existence of meaning. “Always keep Ithaca in your mind / To arrive there is your destiny / But never hurry the journey,” wrote the poet C. P. Cavafis. But then he recalled why one must rein in the eagerness for results: “Ithaca gave you a beautiful journey / Without her you would never have set out / But she has nothing left to give you.” It is possible that the way AI annihilates us is by answering our deepest questions with just a keystroke on the intro.
Read more The X1 bus will go to the Zona Universitària of Barcelona starting September 21
