Making Art With Artificial Intelligence (DALL-E, NightCafe, StarryAI)

When I first heard about artificial intelligence (AI), it was in the realm of a robot-led new world order a la HAL 9000 or Skynet.

Today, it’s become more synonymous with NFTs and machine learning. The level of sophistication that we have gotten to is insane. We have computer scripts that have learned to write a movie scene and create algorithms. An AI even replicated Anthony Bourdain’s voice in the 2021 documentary, Roadrunner, to have his voice narrate an email he had written but never recorded himself speaking. Now, I don’t think AI can create the moving emotion of art or poetry (maybe that’s me falling into the hubris of humanity) but I am rather impressed with its current capabilities, which I also understand is still in its early infancy compared to where it can go in the future.

I first caught wind of this wave of AI art on Twitter, following a few accounts that posted their creations daily. I was amused by the impressive renderings (and their obsession with adding Homer Simpson to all of their queries) and then my dad sent me a link that he had been following them as well.

Curious, I tapped the first link – a free-to-use AI image creator that I would find had fine, but limited, capabilities. Still, it was fun for a little bit and let me nerd out by combining things like Xenomorphs playing the bongos and Super Mario as a sushi chef. The weirder the search, the better.

Then I noticed DALL-E 2 was available via an invitation. I assumed the criteria was pretty selective – digital artists, photographers, people who can at least pretend to know what they’re doing in Photoshop – but I submitted my name and email anyways. A week or so later and I was invited in.

You log in with 50 credits, one of which is used for each query. You can buy more or earn 15 for free each month. I spent a few of my first credits replicating my searches on the original app to see if DALL-E 2 was any better. Short answer – it is.

I texted out the best pictures and even tried designing a logo for this blog. Below are some of my favorites.

Now, DALL-E was the first one that I became familiar with, but I suspected there were others like it. I did a quick search on Google and came across NightCafe and StarryAI. My experience with them is much less, but so far they seem more limited, mostly because they have you pick a specific style of art before the image is generated. I’d rather type in “Surrealist picture of…” than click a thumbnail of what I feel the picture will end up looking too much like. It just seems more cookie-cutter to me, but maybe that’s just me.

I fully expect AI to continue to show advancements and become more integrated in our everyday lives. The rate we’re going, I wouldn’t be surprised if it is sooner than later, that our daily choices are heavily influenced by AI. What do you think?

Wanna try? I’ve included links below.