They do, for even stranger reasons: " In particular, memory functions must be vastly non-lossy, otherwise retrieving them repeatedly would cause them to gradually decay".

To paraphrase babbage, "I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a statement."

Being lossy does not imply reading is lossy, it could be that writing destroys old information.

How funny, retrieving memory in human brains do expose them to potential decay or alteration.

http://phenomena.nationalgeographic.com/2013/05/20/when-memo...

> Every time we bring back an old memory, we run the risk of changing it. It’s more like opening a document on a computer – the old information enters a surprisingly vulnerable state when it can be edited, overwritten, or even deleted. It takes a while for the memory to become strengthened anew, through a process called reconsolidation. Memories aren’t just written once, but every time we remember them.

The wording is confusing, but I thought perhaps they were assuming that an artificial brain would indeed be lossless and therefore could not be modeled computationally.