A) It's only a class; probably only a problem set. If you're at a hot
startup, how much does your transcript really matter? Even if you aren't, how
much does your transcript really matter?
B) It's only a startup. If you're unlucky, it'll crash before the end of the
term (and assuming no rights are assigned) you'll be able to use it for the
homework. Reasonable NDA's don't poison information, they merely color it.
If the party with whom you contracted is your sole source of information, then
don't disclose it. If you learn about a subject from third parties (either
before or after), it should be fair game for discussion. (exercise for the
reader: examine the parallels under garbage collection between NDA's and weak
Both of the above arguments can be moot. An assigned problem should be
soluble given information associated with the class itself or its
prerequisites. Sure, it's more interesting to have "open" problems involving
environmental influences, but even if one knows of a much better, or more
elegant solution to an assigned problem, I can't see why one wouldn't get full
marks for a solution solely involving techniques presented in class, and I can
see reasons to be marked down for presenting a solution based little upon the
Universities originated in feudal times, and were strongly influenced by the
church of the day. It's no surprise that their attitudes towards both
intrafaculty and faculty-student roles suffer a bit of a culture gap.
Of course, the educated medieval man would view the gap from the other side.
After all, he would reason, it only takes a certain number of laboratores and
bellatores to attend to worldly matters; a society should be judged by how
well it can support its oratores' concern with higher matters, and
heirarchical organizations. Beware of annoying the bishop; he might require
you to found a college in penance.
Jane Jacobs wrote "Systems of Survival: A Dialogue on the Moral Foundations of
Commerce and Politics", in which she argues that there are two distinct moral
syndromes -- commercial and guardian -- and that while each alone is
self-consistent and exhibits positive behavior, mixing the two leads to
corruption and negative behaviors. In mixing the economic with the academic,
this case seems to fall right into her argument.