Prioritizing Safety Over Usefulness
Renowned as one of the “godfathers” of Artificial Intelligence (AI), Prof Yoshua Bengio has expressed remorse regarding the trajectory and swift advancements in the field.
In hindsight, he acknowledges that had he anticipated the rapid pace of AI’s evolution, he would have placed greater emphasis on safety rather than mere utility. This acknowledgment follows growing concerns expressed by AI experts regarding the potential dangers AI presents to humanity.
The Call for Regulation
Joining the growing chorus for AI regulation, Prof Bengio has emphasized that granting AI powers to militaries should be avoided. He believes that the increasing sophistication and power of AI systems make them vulnerable to exploitation by “bad actors,” including terrorists and individuals with malicious intentions.
Prof Bengio’s concerns reflect the urgent need for robust governance and ethical frameworks to ensure the responsible development and deployment of AI technologies.
Personal Toll and Uncertainty
The weight of these concerns has taken a toll on Prof Bengio personally. As his life’s work, which has given him direction and a sense of identity, faces increasing ambiguity, he admits to feeling lost.
Navigating the ethical challenges and potential dangers of AI has become emotionally challenging for those deeply involved in the field. Nevertheless, he remains determined to engage in discussions, encourage critical thinking, and foster collective responsibility.
Fellow Godfathers’ Concerns
Prof Bengio is not alone in his apprehensions. Dr. Geoffrey Hinton, another esteemed AI “godfather,” has also expressed similar concerns and recently retired from Google, expressing regret for his contributions to AI. However, the third godfather, Prof Yann LeCun, has downplayed apocalyptic warnings, suggesting that such concerns may be exaggerated.
The Need for Accountability and Ethical Training
To mitigate the potential risks associated with AI, Prof Bengio asserts that all companies developing powerful AI products should be registered, monitored, and audited by governments.
Additionally, individuals working closely with AI systems should undergo certification and receive ethical training. Prof Bengio advocates for treating AI development with the same level of scrutiny and regulation as other sectors, such as aviation, automotive, and pharmaceuticals.
Balancing Concerns and Progress
While some share Prof Bengio’s fears about AI’s impact on humanity, others argue that more immediate and tangible issues, such as AI bias, predictive policing, and misinformation spread by chatbots, require immediate attention.
They emphasize the need to focus on addressing these concrete harms rather than fixating on hypothetical risks. Despite the concerns, AI has already demonstrated significant benefits to society, ranging from the discovery of new antibiotics to advancements in neuroprosthetics.