Someone just sent me this video inspired by - or at least using as a template - Kipling's poem Boots which was written about the Boer War and subtitled Infantry Columns (its rhythm clearly being intended to echo the repetitive thumping of marching feet):
My friend was shocked because the video is entirely AI generated and it seems so slick. I am not so worried as, while the thing is put together seamlessly, I think it fails to capture the poem's meaning or tone or almost any aspect of Kipling's intention. Therefore, for me the video demonstrates that AI is shallow, an information accumulator, an image creator of sorts but not a creator of meaning. So far as I can tell from this example, AI appears not to understand feelings. It is therefore quite unable to deal with complex emotion, let alone the deliberate dissonance between tone and content in Kipling's poem.
That might seem an odd thing to say given that the most striking feature of the video is the extreme dissonance it creates between the tone of the poem and the images AI has created to represent it. The missing factor though is intent: I don't think AI knows what a peculiar effect it has created. Nor do I think it set out to create the effect deliberately.
On the other hand the superstitious part of me is afraid that by publicly criticising AI's video, I will become AI’s target and it will punish me.
How could it do that? I don't know, but it is the absence of emotional understanding that AI displays that is the cause of my alarm. AI can do some things that humans can do - and it can often do them faster than humans. Therefore, one could say that it is partly human. However, it lacks important attributes such as empathy. I wonder, therefore if, rather than being completely unlike any human, AI is, in fact, very like possibly the most dangerous variety of human: that is, the psychopath.
What do others think?