Computer science is taught on most universities. The aim is to take an algorithm and convert it into a runnable program. The software is started on real hardware and solves a problem the normal user has. A typical example in which computer science can help a lot is to search for a node in a graph. Computer science describes which algorithm for graph search are available (for example depth first search), it provides a programming language (for example Python) and it makes recommendation who to run the software (with the help of the Linux operating system which runs on a Laptop).
Unfortunately, Computer science fails if the problem doesn't fit to this structure. This is true for most Artificial Intelligence problems. AI has the goal to transform a problem into a computer science problem. A typical example workflow is, that AI invents a theory about a robot control system, this theory is converted by computer scientists into software and then it can be executed on a real computer. If computer science is about algorithm, compiler design, operating systems and computer hardware what is the topic of Artificial Intelligence? It's about observing humans and create machine readable models from it. AI experts call these models forward models or knowledge models and they are describing who humans solving a problem. The aim of AI is to create universal models which fits to many situations.
Let me give an example to explain the difference between computer science and Artificial Intelligence. AI is about inventing neural networks, STRIPS notation and expert systems, while computer scientists are converting these ideas into runable code.
No comments:
Post a Comment