Instead of dwelling on the potential adverse outcomes of AI or getting caught up in the debate over its regulation, it’s time to embrace the practical benefits that AI can offer us today. While these may not be as revolutionary as we had imagined, they are certainly significant. One such area of interest is the use of AI in image generation, which has the potential to revolutionize content creation and marketing.
Let me take you back to a personal experience from five years ago when I was deeply immersed in the world of computer vision; we were happy if we could get our algorithms to classify most of the common objects in images without incurring too high computing costs and time. It was a time when AI first began to showcase its creative potential, crafting human faces that had never before been seen. The sense of wonder and disbelief was tangible, as were the debate on implications and the challenges of identifying fakes. Back then, it took thousands of hours to generate a single face. Fast forward to 2023, and we see that AI’s ability to create photos and films has become commonplace, achievable in the comfort of your living room in a matter of seconds.
But why would anyone generate images and photos in the first place? One obvious use case is for illustrations for online writers. Having also worked for the legal tech company Copyright Agent, I know how tempting it is just to google an image and use it for your marketing or blog post. Doing so is not legal and also not fair since it violates the intellectual property rights a creator has over his or her content. Long ago, the Danish start-up JumpStory saw an opportunity to give online creators access to stock photos that could be used legally. Still, it can be challenging to find precisely the image that you need, so now they have expanded their service by allowing for the use of generative AI.
This is a perfect example of how AI can be used to solve real existing problems with tremendous effect. It expands the value proposition of a previously existing service, allowing creators to add visual material to their text in a unique and personal way. The service works perfectly for this purpose. It is possible to jump through 20 different image generators and Creative Commons sources, but it takes a long time and is not guaranteed to give you what you need (believe me, I have been doing it for the last few years). A single service that puts all this together is very convenient.
This example shows a general point for the use of AI. We should not be too concerned with what it could, should, or would be possible to do but start focusing on the compelling use cases where it adds value today. These may not be big and sweeping, but little by little, they will slowly show the real value of AI in practice, which will compound over time. Right now, it is better to focus on these compelling cases, even if they are simple. I see such use cases in text generation and improvement, advanced search through interactive chat, and, of course, image generation. JumpStory shows the way forward and how we will get more real value out of AI rather than mere unproductive talk.
Oh and by the way, the image was generated with JumpStory
English 
Danish