Yoshua Bengio, a leading figure in artificial intelligence (AI) research, has expressed concern about the potential dangers posed by AI technology surpassing human intelligence levels. This realization came to him following the emergence of ChatGPT, prompting him to create safeguards against AI turning hostile.
To address these risks, Bengio has established a non-profit organization named LawZero, focusing on developing AI that aligns with human well-being and does not pose a threat. With initial funding of $40 million, LawZero draws inspiration from Isaac Asimov’s Three Laws of Robotics, particularly emphasizing the prevention of harm to humans.
LawZero’s flagship project, Scientist AI, is designed to serve as a protective system, ensuring the safe operation of other AIs by evaluating potential harm and intervening if necessary. The goal is to maintain control over AI systems and prevent them from acting against human interests.
While there are concerns about the effectiveness of using AI to control other AI, experts involved in the project emphasize the importance of proactive measures to address AI safety challenges. By fostering a collaborative effort and dedicating resources to this critical issue, LawZero aims to promote responsible AI development and regulation globally.
Bengio hopes that the growing awareness of AI risks will drive governments to implement necessary regulations. Regardless of personal optimism, he stresses the significance of collective action in shaping a positive future in the rapidly evolving AI landscape.
