123: AI Art Gallery
Here are ten of my original drawings.
Directly next to them are their corresponding machine-generated variations, after I fed them into an AI image generation tool (DALL•E 2).
Can you tell which are human-made?
1/ Falling Cats
From Cats and Toast
2/ Battlestar Galactica
From Battlestar Galactica
3/ A Knight
From Drawing + Writing
4/ Inbox
From MBA - Intro
5/ Meditation
From Past-Future-Present
6/ Easel
From Art for Amateurs
7/ Low-poly Mountains
8/ Camels
From Negotiation - Part I
9/ Subway Scene
From Subway Comic
10/ Turtle
From New Year
Quick thoughts
The underlying technology here is incredibly powerful. And I've barely even scratched the surface re: DALL-E's natural language input capabilities (generating images purely from text descriptions). These emerging AI design tools will only get better, fast.
And yet, a part of me feels trepidation.
According to basically every sci-fi TV show, we should proceed carefully.
I am simultaneously excited, concerned, and curious as to how this will shake out.
What do you think?
Related :
I got to play around with DALL-E this week 🤖
— Jeremy Finch (@thefirejar) May 31, 2022
Quick thoughts:
- The natural language interpretation is good
- I tried to stump it via conflicting inputs, but it was able to reconcile them nicely
- Strange prompts do particularly well
- It's fast
Ten AI-remixed drawings below 👇