AI in the future of graphic design. I have tried many different types of AIS that generate images and experimented with different things that it could help with. The banner image for this post was using an AI. The specific one I used was Dall E 2.

The more I keep watching the industry and seeing the rapid pace at which the AIs are slowly taking over and the speed at which they can run I’m uncertain exactly what the future of craft design will look like. The fact that AI can produce a full-blown high-resolution image and a span of 10 seconds whereas a person can create something similar can take eight to 18 hours or longer to draw by hand or even make it in 3d. Humans have next to no chance to compete with these artificial intelligence systems.

I think they are a double-edged sword, the same thing with the internet: they can be both good and bad. Like in my case I am using it for a banner image. If I were to make one myself it would take many hours which I do not have so doing that is helping me. And it’s also free for the most part. 

 If people are using it for not good intentions like trying to make political images or trying to make the images seem more expensive than they are then I think we’re going to be seeing problems in our industry. 

A large problem I’ve been seeing with text-to-image AIs is the use of NFTs or people trying to pass off the text to AI artwork as their artwork. The main thing with that is the rapid piece of these AIS can create images but they are still not focused on the topic. Because when you create any one of these AI images the computer has a hard time focusing on certain aspects that are quickly changing. There was an update and Dall E 2 where people can put focus on people’s faces that did not exist before. 

So basically people were able to insert their face onto a rendering and have that rendering be in a different style that type of thing the AI is struggling with just several months ago and now they perfected it. I think that the problem too is that the rapid piece of these AIS can outstrip the speed at which humans can make stuff. If someone were to trade to take a person’s face and insert it into an image and try to keep it the same style is very difficult to do. These new systems can do this in a matter of seconds just by having a large sample of numerous images with the same type of style.

The other problem that I’m seeing is that people are trying to use artificial intelligence to create images and trying to sell them for profit as NFTs. Personally the NFTs or practically worthless there are very very few that have any sort of value. People can just right-click and copy them they have next to no value.

Personally, as a graphic designer if I were to want an image that’s that exact resolution I could care less if it’s a copy it does not matter if it’s the original or not to what the same type of image if there were some sort of thing preventing from copying it still would not matter someone can take a screenshot the whole the aspect of entities are garbage. The only thing I can see them being useful for is keeping track of items in-game environments like having items being copied from one place to another and keeping track of what is what. That is the only thing I can think of that would be relevant to that

So the future of text-to-image generation through AIS I think you’re way up in the air right now they’re very very new and they’re to the point where they’re progressing and breakneck piece so it will either do one of two things I can think of he’ll either help the art industry by reducing the amount of time it takes to make artwork. Or it will cheapen the art industry like the art Mills did way back in the early 2000s in the 2010s. The art Mills crowdsourced a lot of artwork so all of a sudden designing and artwork you’re having thousands of people looking at one thing trying to fight for money. In this process, people are paying less and less for artwork and people just thought that they could go and pay for a logo for $5 even to this day this mentality still exists when reality a $5 logo should be roughly $200 if it’s professionally done.

I think this is the same process except instead of having a thousand people trying to make it image or logo you’re having a thousand processors doing the same thing only difference right now is that the AIS cannot generate coherent text from icing I think they will be able to but at the moment it’s coming out is gibberish I have tried on multiple different types but they are strictly more image beast than design beast right now. They can generate things like logos and icons fairly easily. icons don’t usually have text but if someone added in later work for things like that this type of process will work perfectly. But this whole thing is completely up in the air and I think it’s going to be a new evolution for the graphic design field