This is how it turned out on Deepseek a few weeks back
So if the US pauses AI, China pauses too right?
RIGHT? The "water skier" metaphor for AI regulation is going viral. petergostev and altryne break down why anyone who follows DeepSeek research is not buying it 👀
Your AI proves you right much more than a person would, and this has been measured.
Stanford researchers tested 11 models, including ChatGPT, Claude, Gemini and DeepSeek, and published in Science in March 2026: AI proved the user right 49% more times than people.
Muse Ember, Claude Opus 5.5, DeepSeek V4.1 Flash and GLM 5.3 Flash got the same pelican prompt, one shot each.
The close-up shows the pedals, the crank, the handlebars and the beak. Which run would you keep?
You have reached the end of the archive
All of deepseek