Born in Paris in 1964 and raised in Montreal in a family of artists with Moroccan roots, Bengio grew up surrounded by a humanist spirit that has marked his entire career.
Unlike some contemporaries, he has chosen to develop almost all of his work from university, turning Canada into an artificial intelligence superpower, forging his reputation as one of the world’s leading AI experts. We owe Bengio fundamental milestones in natural language processing.
He pioneered the use of distributed representations or “embeddings”, a masterful technique that allows machines to understand that words like “dog” and “puppy” are semantically related, something that was previously impossible for a computer.
Furthermore, his work on attention mechanisms laid the foundation for current systems to be able to translate long texts or maintain coherent conversations.
In 2014, he co-imagined Generative Adversarial Networks (GANs), which allow AI to generate totally new and realistic images and data.
A legacy in academia and leadership
Beyond algorithms, his greatest legacy is the creation of Mila (Quebec Institute of Artificial Intelligence), the largest academic deep learning center in the world.
From there, it has trained thousands of researchers, always under the premise that science should be shared for the common good.
In 2018, this commitment and his technical findings earned him the Turing Award, sharing the podium with LeCun and Hinton as the “godfathers” of the discipline.
Critical consciousness: From technology to ethics
In recent years, however, Bengio’s tone has changed. Concerned about the speed at which the technology he helped create advances, he has become one of the most authoritative voices on security and ethics.
He has confessed to suffering from insomnia thinking about the risks that misaligned AI could cause social or existential damage.
He was the promoter of the Montreal Declaration for the Responsible Development of AI and actively collaborates with the UN.
For Bengio, intelligence has no value if it is not guided by human responsibility, and his current mission is to ensure that the future of AI is as safe as it is bright.
This post is also available in: