Yoshua Bengio warns about AI, says focus on safety, not speed

Yoshua Bengio, a leading AI pioneer, has expressed concerns about AI's rapid growth, advocates for safety over speed, ethics training, and responsible development to counter misuse.

Yoshua Bengio expresses concerns over current state of AI development. (Pixabay)

Renowned computer scientist Yoshua Bengio, a pivotal figure in the advancement of artificial intelligence (AI), has shared his reservations regarding the breakneck pace and inflated expectations surrounding AI. In an illuminating conversation, Bengio admitted that, had he foreseen the rapid evolution of AI, he would have placed a greater emphasis on safety over sheer functionality.

Navigating Uncertainty: Bengio's Call for Collective Progress

Bengio stressed the urgency of pushing forward, engaging in thoughtful discourse, and fostering a collaborative spirit. Recently, he joined forces with other influential figures in the industry to co-sign an open letter that sounds the alarm about potential risks tied to AI's rapid growth and proliferation without any apparent safeguards.

Reflecting on this in an interview with Al Jazeera, Bengio said. "One might say I feel adrift, but the imperative remains to persevere, to exchange ideas, and to rally others to join the intellectual endeavor." The remarks from the esteemed Canadian computer scientist come in the wake of his involvement in a thought-provoking open letter penned by industry leaders, underscoring the perceived "threat of extinction" to humanity posed by AI.

AI's Trajectory Under the Influence of "Negative Actors"

Bengio's present and pressing concern revolves around the misuse of AI by what he terms as "negative actors." These actors could range from military entities to terrorists, or even individuals driven by extreme anger or psychosis. He underscored the peril of AI systems being susceptible to manipulation for harmful purposes, whether orchestrated by military factions, terrorist groups, or individuals with nefarious intent. To counteract this, Bengio proposed a model of government regulation for AI development, akin to the oversight frameworks governing sectors like aviation, auto, and pharmaceuticals.

Upholding Ethical Fortitude

As a countermeasure, Bengio introduced the notion of instituting ethical training and certification for professionals engaged in AI system development. He championed the cause of making ethical training a standard norm within the repertoire of computer scientists. While Bengio and his contemporaries, such as Geoffrey Hinton, convey deep-seated concerns about the existential threats woven into AI, Yann LeCun, another luminary in the AI landscape, has adopted a divergent stance—opting not to endorse the recent risk statement. This disparity in viewpoints signals an ongoing discourse within the realm, centered around the evaluation and management of potential perils.

In the face of these intricacies, Bengio maintains a hopeful perspective, asserting that collective endeavors can surmount the challenges at hand. Drawing an analogy to the fight against climate change, he posits that, even though one cannot undo past actions, the opportunity to effect positive change is never beyond reach.