3️⃣ AI Ethics with Lauren
Finally, I get to write about memes along with ethics! I am taking a deep breath of relief as I haven’t heard of any major nefarious uses for the open DALL-E Mini just yet, aside from a bit of inevitable dark Reddit humor. Rather, like many of you, I have been trying my best to hold my composure over the absolute hilarity of the search results turned memes plastered over the Internet. DALL-E Mini still shows an incredible level of construction and fair detail, and the incorrect rendering adds to the comedic impact of the image, akin to deep frying. The blurriness takes some of the concern away from hyperrealism leading to misinformation, but still gives users all the fun of generating images with silly prompts. I’d say it’s a fairly safe release to the public due to this obscuring feature. We can see this from Louis’ video comparing DALL-E 2 to DALL-E Mini, where the quality of the former strikingly outweighs the latter, and I stand behind not releasing the full DALL-E 2 - no matter how good the memes would be. Again, part of the humor comes from the error in the images!
As DALL-E Mini cannot self-generate images, I am not too worried about the offloading of creativity from this development. I don't think this will put graphic designers, meme creators, or any visual creatives out of business. Rather, this has become an incredible tool for enhancing creativity, to see multiple iterations of whatever ridiculous prompt comes to mind in a strange, somewhat blurred rendering.
If you’re wondering what my favorite DALL-E Mini generated meme is, so far the winner is this one.
Happy AI meme-ing!
- AI Ethics segment by Lauren Keegan