If you take a look at the sources, most of those are "leftist" sources, to whom Elon Musk is a villain. Meaning, the AI is really just a reflection of data created by humans. Now it would be interesting to know why those especially have been chosen but that's outside of this scope.

I'm pretty sure Elon knows about this issue and wants to fix it, but I doubt he'd put it offline, as that would go against everything he publicly stands for. Rather, he'll probably will try to get the AI training data as optimal, unopinitionated and truth-seeking as possible, which clearly is difficult in todays political landscape.

Here's a quick example of my viewpoint:

Given that you're building guitars, imaging building one with poor quality ingredients. Regardless how much effort you put into it, the output will be less good as if you're using the best resources available.

And I'd argue the same can be said about AI and data. If the data is poor, the output will be poor. If the data is optimal, the output will be optimal. This is especially true as AI is getting trained more and more on data created by AI.

So in the end, truth (aka good data) really matters to get an optimal AI.