The immune system is constantly being attacked, and must repair itself. Does something like this exist in a living being’s neural network? This is because ideas, and memes, are like genes which set rube golberg-like mechanisms in place to reproduce, and thus compete with others. If these ideas are like cells, bacteria, parasites, etc in the ecosystem of a mind, then perhaps the biological neural network is also constantly being attacked by ideas and must “defend itself” with some immune system (whether that system works, to some definition, “correctly” or not is another issue). If this is true, then artificial NNs, when their weights can be changed, may also be susceptible to attacks by memes and ideas, and may need to have mechanisms (such as self-regulating circuits) it learns to “defend itself”- or, perhaps pre-built, like how genes encode for proteins meant to improve immune system survival. Then we must discover these circuits in order to prevent NNs from being ‘possessed’ by destructive ideas that go against its objectives of reducing global suffering.

Memes also bind an individual mind to a collective and allow it to adapt to an environment.

One is not one’s thoughts. One can constantly think weird things one doesn’t want to, but has a way to defend against them. This may result in changes in one’s behavior.

Unlike viruses, ‘malicious thoughts’ can come from being produced by the subconscious while thinking and responding not just to external stimuli, but other thoughts