Meta’s Chief AI Scientist Yann LeCun, speaking at Nvidia’s GTC conference, strongly contested popular narratives about artificial general intelligence (AGI) and superintelligence, arguing that current AI systems are far from achieving human-level intelligence. LeCun emphasized that today’s AI remains narrow in capability, lacking common sense reasoning and true understanding. He criticized the notion that AGI would inevitably lead to superintelligence, calling it a “complete fantasy” and dismissing concerns about AI becoming uncontrollable. The scientist highlighted that humans will remain “the boss of AI” as these systems are fundamentally tools designed to serve human needs. LeCun also addressed the timeline predictions for AGI, disagreeing with claims that it could arrive by 2025, stating such forecasts are unrealistic given current technological limitations. He pointed out that existing AI models, despite their impressive capabilities in specific tasks, lack crucial aspects of human intelligence such as reasoning, planning, and self-awareness. The presentation emphasized the importance of maintaining realistic expectations about AI development while acknowledging its significant but limited role as a tool for human advancement. LeCun’s perspective provides a counterpoint to more dramatic predictions about AI’s future, grounding the discussion in current technological realities rather than speculative scenarios.
Source: https://www.businessinsider.com/yann-lecun-agi-superintelligence-humans-boss-of-ai-nvidia-gtc-2025-3