Nature has an excellent track record in solving problems, and while biological inspired approaches draw inspiration from nature, they should not emulate it blindly. What works for nature may not work for computer systems - bio-inspired computing comes to the rescue. In this position paper, we look into the problem of bio-inspired computing from two perspectives, that of models and algorithms. In the context of self-adaptive software systems, the challenge is to come up with approaches that are able to generate specific solutions on demand and during operational-time.