In Python 3.x, byte strings and Unicode strings are separate data types, and it's important to be explicit when working with bytes to avoid encoding errors.

But you could relax this requirement by specifying a more forgiving error handler (see Python's Error Handlers docs for more options): ...

lu = rpc.listunspent () >>>lu [-2] {u'account': u'', u'address': u'2NCgTCaiUHSVKZ6Wwvx74D1Me9gLN1Gw8Rr', u'amount': Decimal ('0.00940000'), u'confirmations': 14839, u'redeemScript': ...