Great article George and I agree with the statements to some extent. However I would like to point out two issues here:
1 - In some fields such as credit scoring or legal decision making, the logic behind the algorithmic decision making system (i.e., AI system) must be explainable. This is a requisite of the law and we must have AI systems which are both explainable and very accurate simultaneously;
2- Assuming that the human brain has the most sophisticated structure ever existed and ever will be is a bit narcissistic. We may come across to AI algorithms that works differently than human mind and have a better judgement. Therefore, although the currently the best performing models are the ones which mimics the neural networks, there is no guarantee that we won’t find a strange algorithms which outperforms everything. So, in sum, neuroscience may not give us the ‘ultimate answers’ in AI field.