Currently, quantum computers have a limited number of qubits (quantum bits), which are the building blocks of quantum computation. To perform complex AI tasks, such as training deep neural networks, a large number of qubits would be required. Achieving the necessary scale and stability for quantum computers is an ongoing research effort.
Traditional AI techniques, such as deep learning, have been highly successful and are well-suited to classical computing architectures. It will take time and significant research to develop quantum algorithms that outperform classical AI algorithms for a wide range of tasks.
Classical computing systems, on the other hand, are readily available, scalable, and cost-effective for most AI tasks.
As the technology matures and quantum algorithms improve, we may see more integration between quantum computing and AI, but it will likely involve hybrid approaches and a combination of classical and quantum systems.