# Generative AI Explorations

By [Ashish Kumar](https://paragraph.com/@ashventuredoteth) · 2024-04-24

---

Generative AI's journey began with foundational research, when the groundwork was laid in the development of Generative Adversarial Networks (GANs) around 2014. These algorithms act like artistic rivals, one creating images, the other critiquing them, leading to a continual refinement process.

This technology exploded into the public eye in 2022 with the launches of platforms like OpenAI's [DALL-E](https://openai.com/index/dall-e-3/), [Midjourney](https://www.midjourney.com/showcase), [Stable Diffusion](https://stability.ai/news/stable-diffusion-3) and Adobe's [Firefly](https://www.adobe.com/in/products/firefly.html) (public beta in March 2023). These user-friendly interfaces allowed anyone to describe an image, and generative AI would conjure it into existence.

In 2023, Adobe Firefly's launch brought a revolutionary development to the world of image editing. Unlike other AI tools, Firefly seamlessly extends existing images within Photoshop, without requiring any prompts! Intrigued by this capability, I pushed its limits by feeding it classic art. The results were incredible, sparking a fascination with Adobe's Generative AI.

![](https://storage.googleapis.com/papyrus_images/0d5057ff7d47d8b6724f5e6174b303993524fe657579b43ae1f45a0e22c75efe.jpg)

![](https://storage.googleapis.com/papyrus_images/9f57aa0ddbcd64ffcac3926d86f39eba6767ca9c068f0725e8d43bf9096f094c.jpg)

![](https://storage.googleapis.com/papyrus_images/ecb7ab56bb754ba53d402c6fa39cbbfc484ac5fa05eadb8d79835567bd1d872a.jpg)

![](https://storage.googleapis.com/papyrus_images/d13dbbeaa188260bfb3296ff5ea1bcdcfa57ddbd59887631acfc2f3de269f770.jpg)

![](https://storage.googleapis.com/papyrus_images/b97826144cc4faa7c8060024d56710cfbf2c4b3e1d747f960879c7b2e888ff67.jpg)

![](https://storage.googleapis.com/papyrus_images/9de6ffe842c2f1d17f2aa02e3cc5bbef65761b5d00ee852d67a0455701646ce1.jpg)

![](https://storage.googleapis.com/papyrus_images/974e9c49e52c82b2be84b74ff8f58a5796a4464ed74fd542b4d51da164b90824.jpg)

![](https://storage.googleapis.com/papyrus_images/0fe78f691c2865c2747f11f816dd302fe6c95ebbaa3dc88f733ad0dc1e66d87c.jpg)

![](https://storage.googleapis.com/papyrus_images/451db1a3bfe8443927d0df901d939b16619554b5a7c1bb87171960a7b24409bb.jpg)

![](https://storage.googleapis.com/papyrus_images/c768315d44132dbaa641119fc451619763d4d4e09dd6575cdb5cb31c950fcedd.jpg)

[https://x.com/ashventure/status/1666512709129498625](https://x.com/ashventure/status/1666512709129498625)

Since then, Adobe's Generative AI capabilities have grown tremendously. Today, image fills would produce even more impressive results than what I achieved back in 2023.

### Midjourney

I subscribed to Midjourney in the middle of 2023 and have been using it to generate images for brandwork and refine my prompt writing ability. I also took a [prompt engineering course](https://maven.com/nick-st-pierre/midjourney-for-creatives) by [Nick St. Pierre](https://twitter.com/nickfloats).

![Some of My Midjourney Explorations](https://storage.googleapis.com/papyrus_images/ef504cb14c556a9cb484eabdebe242dfe8dbb062eae13e62db4906919ada6296.png)

Some of My Midjourney Explorations

---

*Originally published on [Ashish Kumar](https://paragraph.com/@ashventuredoteth/generative-ai-explorations)*