so if you have not yet put it together, your data is being fed into a palantir language model
so later they can ask a faulty model which people are likely to be terrorists, or communists, it will certainly become more arbitrary as it progresses
and there is a straight line to the recently used faulty targeting models used by the isreali forces to justify targeting civilians in palestine
we only need to use the rest of our resources to do it, and this isn't up for a vote