DALL·E mini is a free online artificial intelligence(AI) model that can generate images from any word input you type. Since the AI model is in it’s infancy and still learning, you’ll have to experiment with different word inputs to get results closer to what you’re looking for. You’ll also have to imagine how this technology will develop in the near future.
After experimenting, I quickly realized that you have to input more recognizable names in order to get anything resembling a likeness. After a series of disappointing test runs, I saw these results someone else obtained by inputting “Bruce Willis eating a yogurt”.
So, I typed in “Arnold Schwarzenegger eating a yogurt” and got reasonable results, given what you can expect at the current time. As a result, I gave it a “pass”. Keep in mind, this technology will blow your mind in a few years.
I switched up the menu:
Lou Ferrigno’s meal arrives:
“Sergio Oliva eating a yogurt” wasn’t on par with the Schwarzenegger or Ferrigno results. There are a lot less images floating around of Oliva for AI to learn from compared to the amount of Ferrigno or Schwarzenegger photos available.
You get the idea. As hard as it is to believe based on these results, you’ll see this technology continue to improve to a quite impressive level. Give it a try: https://huggingface.co/spaces/dalle-mini/dalle-mini