The beginning of last year was eventful in that we realized what modern AI, Dall-E can do to revolutionize how humans do art. Quickly there were many articles, videos, explainers and social media posts about how amazing or dangerous this new technology is. No matter which group you are in artificial intelligence-powered tools are not going anywhere in fact, they will continue to improve with unprecedented speed.
![](https://miro.medium.com/v2/resize:fit:700/1*039bw6ASWAQsXfUcaOdFjw.png)
![](https://miro.medium.com/v2/resize:fit:1024/1*X-fFqWTDCDXO_AGFYuHm0w.jpeg)
![](https://miro.medium.com/v2/resize:fit:1024/1*fA-7Ka9UIebbD4Ry2sjhfA.jpeg)
Towards the end of the year, we saw another AI innovation that had the same effects on the public eye even more so dramatically. ChatGPT proved that AI can creatively answer complex questions and create articles, essays, scripts and even whole stories. People who wanted to utilize this technology realized that this tool can be added to their workflow after extensive experimentation. There is no doubt next year will also be a very eventful year for AI innovations considering the exponential growth of this technology.
At the current stage of AI, we can safely assume that Dall-E and Midjourney will perfect their algorithms and become the go-to tool for art creation. ChatGPT on the other hand possibly will be the most used AI text generation tool.
The one technology that is needed and does not yet exist is AI-powered 3D model generators. Our team is working to fill this gap by delivering a service that can create 3D avatars quickly using AI. This problem, however, is more challenging than 2D image creation (Dall-E) considering the 3D models have more complex surfaces. Nevertheless, because the textures on 3D models are usually saved as 2D images with enough time given it is possible to train the AI optimized for 3D avatar textures and bodies.
At this stage of our software (Cymorph 3D), we can easily turn 2D selfies into 3D realistic models. The AI simply identifies the eyes, nose, cheeks and all other facial structures from the image then creates a texture file and paints the 3D model using the shades, shadows and many more details from the picture. As a result, we can get very realistic output. Of course, this won’t be the only thing the software does. We are planning to add more and more functionalities to the algorithm.
![](https://miro.medium.com/v2/resize:fit:700/1*Lv5Vsru6yJ8LWGVevKbTWw.png)
The Cymorph 3D will not only be the one-stop solution for avatar creation from the premade assets, but we can also utilize these new AI technologies as GPT3 to make an AI engine that can create the models from a text description. The utility of such a tool would be vast.
We strongly believe the year 2023 will be the year of AI-based tools. We are very excited to develop our software in these times.
This and all the roadmap plans we have will be discussed in our next article in great detail. Follow us on social media and here on Medium to read it when it comes out.
Website: https://cymorph.com/
Twitter: https://twitter.com/Cymorph3D
Discord: https://discord.gg/nSaPxVHXkJ
Medium: http://cymorph3d.medium.com/
YouTube: https://www.youtube.com/@cymorph
Email: team@cymorph.com