A new study published in Nature reveals that AI models can now generate highly realistic synthetic images that are indistinguishable from real photographs to the human eye. The research, conducted by a team at Stanford University, developed a novel neural network architecture that significantly improves the coherence and detail of generated visuals. The technology has potential applications in film, gaming, and design, but also raises significant concerns about the creation of deepfakes and misinformation. The authors call for the development of robust detection tools and ethical guidelines to govern the use of such powerful generative models. Read the full article at https://technologyreview.com/ai-image-synthesis-breakthrough.