Deep Learning Applications
If you have decided to work through the deep learning block, most likely that means that you have seen some awesome deep learning projects and were inspired. This section is designed to keep this inspiration alive. We will not list all the available application of deep learning, rather we will show a couple of the most exicing developments.
Large Language Models
So called large language models have dominated the deep learning landscape for the better half of a decade. We will have dedicated sections to this topic, but let's at least shortly mention one of the newest large language models: ChapGPT by OpenAI.
We will not discuss the intricacies of the model just yet, but let us give you a fascinating example. We asked the model to define the term "artificial intelligence" in the style of William Shakespeare and ChatGPT provided the following poem.
And what happens if we ask ChatGPT to define AI using a poem by Rudyard Kipling?
There are many more things you can do with GhatGPT. You can use it to write and improve code, use it as an alternative to search engines and generate blog posts. While ChatGPT is not a silver bullet, there is no denying that there are tremendous advantages in utilizing large language models. And they are only going to get better over time.
Text-To-Image Models
Text-To-Image models have become extremely popular in the year 2022 with the release of models like DALLĀ·E 2 and Stable Diffusion. As the name suggests, you provide the deep learning model with a text prompt and the model returns a fitting image. The images below were generated using Stable Diffusion.
While the quality of the images is already quite impressive, the future will most likely include models, that are able to generate whole videos and even movies.