Why is it difficult?

The role of a computing machine is to solve a range of optimization problems for specific tasks. While classical computers are designed to solve a large set of general-purpose problems, quantum computers can only solve a handful of problems efficiently due to their underlying physics and probabilistic nature. Therefore, the challenge is to formulate problems in a way that a quantum machine can solve. It requires developing new ways to think about computational problems and re-frame them in the correct format for a quantum computer. It’s explicitly challenging for specific domains, such as machine learning optimization problems, which require a large input dataset. Today’s quantum computers are not suitable for such computation tasks.

What is the impact?

Quantum computers can solve specific problems in split seconds but would take even longer than classical computing for certain other problems. So, formulating an optimization problem in the correct format would greatly benefit from the fast computation of quantum computers. Also, these quantum machines would enable new domains such as machine learning.