Dall-E experiments: my first experience
I’ve got beta testing access to Dall-E! Here is what have I learned from using it
Dall-E is the most expected and well-known AI image creator. Basically, projects like Dall-E have the potential to replace photo and illustration stock services and, possibly, even an illustrator as a profession. Sure thing, it’s more of a hyperbola, and non of the existing AI nowadays can replace any creative human being. But this attempt is worth highlighting.
Disclaimer: This article is not an ad, nor a promotion. This article represents a personal experience using a tool from a design perspective.
Short background
I have heard about Dell-E here and there, famous design publishers showed samples with a noticeable layer of expectations for that almost ‘revolutionary’ tool. Recently, I was interested in the neural network artworks and possible transition them into an NFT space, so I was expecting Dall-E in order to try it out. And I‘ve got an invite!
Dall-E offers the first 50 credits for free, and each credit request contains 4 different image options, which means you get the first 200 (!) images for free. After this limit will be reached, you can buy more credits, for example, for $15 you can get 115 credits more.
What does Dall-E offer?
As I mentioned above, this AI-powered tool creates artwork based on the request in a human language. For example, if I type in “Best picture of Dall-E” it shows the following:
Let’s dive deeper into it. The first picture imitates the Balkan mountains. It seems like a skydiver have just landed right into the trees. Let’s help them 🪂💥 There is also some rubbish and an unknown animal.
The other examples are containing dirty rivers, strange houses, and to me more similar to what you can see diving through Albania with its colorful small houses on hills with unusual roofs on top of each.
Let’s try another request: “a baby girl is asking for Peppa pig”:
Another request: “next generation online events platform for all”:
The last pearl: “attractive housewife making a morning cup of coffee in a hotel room”:
Would you like to drink some coffee from the iron? I think this is a good life hack because usually, you can not find a kettle in a hotel room, but it’s easier to ask for … an iron and warm-up water for coffee ☕️ Genius!
Will AI replace people with real intelligence?
Dall-E is a good attempt of using AI to cover people’s needs. At the same time, it clearly shows that for artificial intelligence, even the smartest, and with the deepest machine learning algorithms (and other ‘cool’ features) is almost impossible to replace real people’s intelligence. At least not in the far future. Nowadays, AI can not provide the necessary content in an expected high-fidelity manner, especially on a commercial basis. The same applies to the content that solves the real business problem: it's beyond AI’s proficiency.
But, this attempt with Dall-E simultaneously has a positive impact: in a concise period of time you can get a 100% legal, unique, non-fungible ‘art’. Sure, the quality of that art is somewhere in between a kid’s daub and low-fi photoshop skills. And this edge of attempts to deliver something in between happens quite successfully. As the output, you can always get something fun, unexpected, and weird that pushes you to think about it. Which usually meets the original goal of modern art: to make you think. With AI you can occasionally get gallery-level artwork, that could potentially be fine art as a result and maybe even be sold for a significant figure. Occasionally.
Can AI create an independent direction of arts?
A possibility window was opened even earlier. Internet was rumored about the neural networks for years. Art made by some code, algorithms, and AI opens up new possibilities with the new page to the book of arts. And this type of art creates a new niche, with a semi-digital world outlook, that bumped into our life with metaverses and XR-tech that imitates and augment our reality with unreal objects and fresh meanings. Altogether, these digital new arts would be a good addition to the hybrid technogenic future.
