In data science and AI, there is a tendency to use off-the-shelf algorithms. However, for innovation, one has to create new algorithms or has to tweak the existing algorithms. It presupposes knowledge of mathematics. You just cannot do it by mere programming.

ASIC or application specific integrated design is an area that has well-proven, established algorithms. Data science still has not reached this stage. Here applied maths is still emerging.

The area of a triangle can be calculated using a well-established formula or algorithm. However, if we have to calculate an irregular area with several curves and line boundaries, this area would require subdivision in such a way that computing is possible. Area could be subdivided many ways, and each throws up an algorithm. However, one solution could be the most efficient one.

In AI and data science, there are many possibilities across diverse industries. This calls for applied maths skills and not computer science skills. A data scientist from programming background can survive for a few years, but faces issues later.

Data science courses should be under the maths department, rather than computer science department.