NJX-njx PRO

you are right

This article is very inspiring to me.

1. Since skills have become a great boost to the improvement of model capabilities, can we try to distill skills, just like we did model distillation before? I think this can be achieved through multiple iterations.
2. The current functions of upskill are actually quite complete, but I wonder if we can try to make it generate a compatibility matrix between multiple skills, so that the combined effect is greater than the sum of the parts. In addition, Model A generates skills, and Model B looks for counterexamples, so that they can evolve together.

Actually, I think a very important point is that most independent developers do not have enough case studies to support their work, and at the same time, the cost of online deployment is actually a bit high

To be honest, although it still looks like AI at first glance in terms of design and other aspects, the quality of the model's website has indeed improved a lot compared to previous models.

So you want to create a dataset of papers that includes various AI-related papers, is that right?

I'm a bit confused. What can such synthetic census data be used for?

Yep! You've got that exactly right. I've finetuned the model so that it when it responds, it's like a Type 1 personality from Enneagram.
great

This sentence is very insightful: In today's world where models are becoming increasingly homogeneous, these decisions mark a significant shift in the competitive landscape from model performance to system design. What China has made open-source has never been a single model weight, but the entire AI ecosystem.

I would love to hear your views on China's system design.

As a Chinese AI researcher who has fully experienced the wave of AI technologies and products in 2025, looking back on the series of changes brought about starting from DeepSeek, I have mixed feelings.

The author's analysis is excellent. The open-sourcing of the R1 model weights and related technologies empowered the most usable technologies to both startups and large companies at that time. Such an "unconventional" change pushed China's AI field, whether in research and development or product development, into a stage of rapid development.

Thanks to open source. May AI technology benefit all of humanity.