Generative Models and the Dance of Noise and Structure

I like to think about what the inhabitants of Renaissance Italy, aflame with passion for the possibilities of the human imagination and rationality, would have found most astounding about our present-day technology. <a href="https://medium.datadriveninvestor.com/there-is-no-clash-between-art-and-science-3028d0420fbe" rel="noopener ugc nofollow" target="_blank">Leonardo da Vinci</a>, dreaming of flying machines, would have surely been impressed by an Airbus 380 soaring through the air, with passengers comfortably reclining in their chairs, watching movies, and complaining about the Wi-Fi not being fast enough. But out of all the technologies that would have seemed like witchcraft in medieval times, the wonders of generative AI might be among the most witchcrafty. What would Leonardo, after laboring endless years on the portrait of the Mona Lisa, have said if I showed him a device that could paint a portrait of a woman in his style in mere seconds? Lo and behold: <a href="https://towardsdatascience.com/generative-models-and-the-dance-of-noise-and-structure-e72fe7494f4f">Read More</a>