So I came across these 2 questions:

 - [Ideas on how to make a neural net learn how to split sequence into sub sequences][1]
 - [Search minimum value with learning machine algorithm][2]

For me both the problems could be easily solved algorithmically (i.e. without any abstraction and having a definite result) without any help of Neural Nets or AI (both of which try to give the best result for a given set of parameters) whatsoever. I also assume that training a NN/AI for such sort of problems will be more time consuming, resource intensive and pointless (I maybe wrong though).

So my question is, if I want to solve a problem, how to decide whether it is better to solve algorithmically (definitively) or using NN's and AI if solution in both the domains exist. Basically the question can be thought of analyzing the pros and cons of the problem in both algorithmic and AI domains and selecting the suitable solution. How can this be done in a systematic way? And if I have to answer someone why I chose that particular domain, what should be my answer?

**Summary:** Choosing between normal computational approach vs abstract approach used in Neural Nets or AI.

Example problems are appreciated :)
Thanks in Advance.

  [1]: https://ai.stackexchange.com/questions/5838/ideas-on-how-to-make-a-neural-net-learn-how-to-split-sequence-into-sub-sequences
  [2]: https://ai.stackexchange.com/questions/4859/search-minimum-value-with-learning-machine-algorithm/4860#4860