the only way for X to win is to prioritize video. they should be aggressively chucking billions at creators to post, build a corpus of video data to train grok. been wrestling with this idea for a while now, whatever i do its always an uphill struggle to get anyone in my social circle that isn't conventionally a 'hobbyist' to sign-up. but the more i think about it its obvious, people are too lazy to read. its less enticing the LEAST efficient way to feed our brain with information. our eyes can do so much more than scan left to right, a picture speaks a thousand words blah blah. but theres another important reason: video data will be the ultimate input for ai models. in the same way that videos have a rich density of information for human brains, the same will be for artificial intelligence. Youtube / google is the company best positioned for this imo and @benthompson argues this well in his article. they have the largest, structured corpus of valuable (short and long form) video content that could train a video ai model, they already have veo3 but imagine being able to take a creators idea from a thought to a published video in 10 mins. not only would that be more efficient but the content would be higher quality. if i were X i'd aggressively allocate a larger creator fund to entice creators to cross-post on X (no extra cost for them) in the same way youtube does.