Act as a machine learning engineer with 5+ years of experience in deep learning. Provide a beginner-friendly introduction to Generative Adversarial Networks (GANs), explaining the core concepts, how they work, and their real-world applications. Focus on the roles of the [generator] and [discriminator] networks, the adversarial training process, and common challenges like [mode collapse] or [training instability]. Include a simple code example using [PyTorch] or [TensorFlow] to illustrate a basic GAN implementation. Tailor the explanation for an audience with basic knowledge of neural networks but no prior exposure to GANs. Use analogies and visuals if helpful.