Neural mining did disappear again, but in this case it was replaced and the new information is 
.
With the help of Cariappa, Amankwah-Crouse found that the G6 DNCs were being used to process close to 85 quintillion (1018) undisclosed databases, all originating from worldwide public and private sources. The filenames showed references to data formats attributed to social media and data platforms, surveillance systems and companies, and financial institutions. Some of the databases overlapped with each other, implying that they had been shared or scraped, and that the DNCs were being trained on partially redundant data.
It goes on from there a bit, but WOW. Seems possible that Xu only pretended to be eco-conscious to get Amankwah-Crouse to work on the DNCs, and instead he trained them on so much illicitly-obtained data that it negated the environmental benefits of the PACOTTI architecture entirely. Then, after Amankwah-Crouse left the project, Xu had G6 secretly retrofitted with quantum neural networks between October and June of the following year.
When that fact came out in a data leak a few years later, Xu didn’t deny it and just claimed it was necessary because of how much data G6 was meant to process…but that’s crap. They overloaded those systems from the start by training them on a bunch of data they never should’ve had. Plus if it was so necessary, it wouldn’t have been done in secret.
What was Xu doing with that data? Could it have been used to train AI on a scale we’ve not seen before? G6 was a pretty collective/collaborative project at that point - theoretically teams from the WHO, Zhupao, and the CCP could’ve had access at least. The thought of that much personal data being so available makes me uneasy.