The Kaylon's biggest failure as an AI species is that they make rash decisions based on limited research, minimal evidence and a serious lack of understanding brought on by their prejudices instilled by the one and only catastrophic experience they've ever had with organics. "This species does this one bad thing, so all species must all do the same bad thing because they are all made of similar organic substances." Now the Kaylon have a reason to worry about all organics wanting to destroy them, but they won't accept or even be aware that it might be due to their recent attempt to exterminate the human species. They'll keep thinking to themselves that the upcoming war against all organics has to be someone else's fault and not their own. In terms of creative writing and what we know about computers today, all of this is perfectly plausible. Just look at how YouTube and Facebook computer algorithms inadvertently push hateful and misleading content to the top of the list simply because it brings in more ad clicks and more revenue to these corporations. The nature of what's being pushed isn't being taken into consideration by these algorithms because there's no awareness whatsoever of the content with the possible exception of sex and nudity, and again regardless of the actual content. What we're seeing with the Kaylon could very well become our own future if we don't program AIs properly.