March 2, 2023 — With the latest and greatest technologies presenting themselves, including OpenAI ChatGPT, in which we developed TrackMyBot, an artifcial intelligence engine that is able to understand your questions and deliver adequate responses in a timely fashion. The technologies of ChatGPT are only expanding and we highly believe they will someday be used in the medical field, the educational field, the autoindustry, the field of business, and so much more. As the technology progresses, the understanding of what it can do is evolving, and so a new technology is forming in the form of artwork. We are happy to introduce to you artsy.sh.
While not expressively called "ChatGPT" in the art world, the technology is similar in that it can take words and draw them into a photograph. The engine technology is called DALLE-2, Stable Diffusion, MidJourney, NightCafe, DeepAI, and many more. Each engine model has its own unique style based on user input and may even produce different results based on what is entered. Some are simply based on keywords, while others use the detailed description of what you want it to draw. This type of technology may even replace the sketch artist so that someone who is trying to describe the face of a criminal can do so to an AI, and in return, receive a perfect image of the person. This can be in the form of a sketch or even in an actual illustration.
This technology is amazing although it will never replace human touch and emotion. It comes very close but humans do not have to feel threatened by artificial intelligence being able to draw as good as a human can. Instead, these images may help those who are less capable of drawing art to produce the ideas in our heads into various artworks. Stock imagery is also in constant demand and this technology may also help produce these images, saving companies hundreds to thousands of dollars a year by being able to draw a vision. Rather than feel threatened, we can embrace the beauty of the future of what computers are able to do.
As a result of the hype around this technology, we wanted to explore it as well and plan to incorporate additional engine models. Currently, we have implemented two included DALLE-2 and Stable Diffusion 2.1. After several dozen drawings between the two, we listed several detailed keywords that can make better photos. We also began to see the highly contrasting artistry perspectives between the two engine models. DALLE is far better at drawing imaginative illustrations, while Stable Diffusion is highly efficient at drawing real life portraits or places. Currently, this project is free to use, with only charges applying if you prefer to use the API in your own web applications.
To get started, simply make sure to pass the captcha while entering either of these two commands depending on the engine model you wish to use.
draw dalle smiley face draw stable house on a hill
Sometimes the resulting drawing will be completely on point and one of the best things you've ever seen, while other times it may be completely off, and may need multiple iterations, sometimes even a more detailed description of what you want it to draw. Illustrations and photos take less than a minute or two to complete. As technology is improved, this time may be reduced, although waiting for a machine to draw your photo is often times worth waiting for. We do hope you enjoy and are happy to bring this technology to you at your fingertips!