$NVIDIA(NVDA.US) has plummeted so badly that confidence seems to be shaken. To be honest, large models' applications in the C-end have already been dismissed—whether it's writing, music, or painting, they're just garbage that fools amateurs. B-end applications are currently just self-entertainment for some leading companies, with no breakout business model in any niche area. Poor applications remain a sword hanging over this wave of large model hype. The AI industry has already experienced no fewer than four cycles of boom and bust: support vector machines, probabilistic graphical models, deep learning, and now LLMs. Each had its moment for a few years, with industry hotspots constantly shifting, forcing many practitioners to follow, whether willingly or not.

Is this just another bubble?