The ratings are currently very high, but of course, I still need to try it myself.
Personally, I think it's not only concise in its responses but also very fast. It rarely gets stuck in repetitive loops and even directly tells me when it might not be able to continue, rather than generating long, endless outputs like the old o1-mini used to.
I feel that both DeepSeek’s open-source approach and OpenAI’s release of the o3 model highlight one thing: the only thing that truly can't be delayed right now is AI learning.
The AI landscape is evolving so rapidly that even career transitions have become highly uncertain. No one knows what things will look like a year from now. That’s why, at this moment, the best approach is to learn as much as possible about AI and gather relevant insights.
Top comments (0)