As the authors say in their paper, "[a] closer look reveals that the source of the problem is in defining learnability as the existence of a learning function r

As the authors say in their paper, "[a] closer look reveals that the source of the problem is in defining learnability as the existence of a learning function rather than the existence of a learning algorithm". Algorithmic decision making under uncertainty is NOT the same thing as finding learning functions over infinite domains. Sometimes such a function does not exist. Why should it? Although the authors seem pretty much aware of this fundamental shortcoming in their analysis, they nevertheless go ahead and publish a whole paper about learning functions! This has a name, intellectual fraud.