Holy crap, that's a lot of unread texts. We definitely should automate translating those. These date back some 8000 years. I've seen some with advanced mathematics and great thinking ... what else could be on them? Certainly a lot of crap (Ahab owes me a sheep) but also certainly a big window to ancient thinking.
100,000 short texts might not be enough to do machine translation, if the parallel texts are even available in a machine-readable form. Wikipedia says these people are trying to post utf-8 texts of 50,000 tablets, so if anyone were going to try it they'd the ones to start with... but they seem to think File Maker Pro's proprietary file format is a good way to exchange data, so they couldn't make it less inviting to play with if they tried.