It is fine to bring the best videos on the internet to X. One of X's core values is cultural commentary.
However, we cannot tolerate big accounts intentionally using their larger following to hijack impressions hours after another user posts something.
Having said that, we will aim to reward original content creators and livestreamers more in the coming weeks.
Grok foundation model V9-Medium (1.5T) has finished training. Evals look good. A lot of Cursor data was added in supplementary training and there is more to come.
Fine-tuning is underway and reinforcement learning begins in a few days. 2 to 3 weeks to public release.
This will be a major improvement over the 0.5T v8-small that currently serves all Grok production traffic, especially for difficult coding tasks.