Robin Li: The Public's Excitement About AI Large Models is Misplaced, I'm Anxious
-
On December 18th, Baidu's Robin Li pointed out that the current wave of large model technology is unprecedented in AI's 70-year history, representing a completely different opportunity.
Li stated: "What makes this different is its generality - it can learn things it hasn't been explicitly taught. With this characteristic, when you have a set of foundational technologies that are very advanced, they can quickly create valuable applications in various scenarios."
Li pointed out that society's focus on large models themselves is wrong. "I see that the main excitement in media, society, and the public is still about foundational models, not shifting to AI-native applications. This makes me somewhat anxious."
In his view, there need to be thousands of AI-native applications built on top of foundational models for the value of large models to be truly realized.
So I keep emphasizing that we must focus on developing AI-native applications. Only by creating these can your models truly have value.