AI will continue to scale up

The tech world is still rattled by the Chinese AI models that crashed the tech market. The AI community is firmly convinced:

Intelligence can be captured by computer models.

Intelligence will scale up, with increasing computer capacity, for the foreseeable future.

Progress is accelerating, and will continue to accelerate beyond our ability to measure it.

Unleashing artificial super intelligence will be the most powerful thing anyone has ever done.

The premise of OpenAI is that it would discover and tame the super-AI, and make it safe for the world to use.

The real threat of the Chinese models is that open source LLMs would pass up the closed source OpenAI and Google models, and leave everyone exposed to whatever good or bad comes from super-AI.

Recent events have not altered any of these opinions. The Chinese models are just what everyone was expecting and fearing. We just didn't know that they had enough Nvidia chips to be players in the AI race.

We will see more and more AI moving to user devices, like phones, cameras, and toasters. But the real action will be in the supercomputers with the most computational capacity. Companies and governments will invest trillions to try to real that super-AI first.