Slowing artificial intelligence down: the alarm, the proposals and the commitments
The risk of losing control of AI worries more and more people, and it is the subject of the moment in the technology community, and beyond it. Here is what is at stake and what the main players are saying.
In March 2023, an open letter from the think tank the Future of Life Institute, signed by figures such as Elon Musk and Steve Wozniak — including the Portuguese professor Luís Moniz Pereira — called for a “pause of at least six months” in the training of generative artificial intelligence (AI) systems. Among the signatories’ concerns was the risk of a “loss of control” over these systems.
More than three years on, the argument about the pace of technological development is back in full force, and now at the highest level. Dario Amodei, co-founder and CEO of Anthropic, the company behind some of the most capable AI models in the world, has published a text in which he declares: “We need to slow down the rate at which we improve the capabilities of AI models.”
AI translation from the Portuguese. First published in ECO on 15 September 2026.