Official 2025 Dall-E Mega Gallery 🍀

Image(Wide, Be humble but creative, NoText())

New images are way too text heavy but my macros are native to 4o now, GPT not required it seems :slight_smile:

1 Like

MotorCycle accident - Ex girlfriend…

2 Likes

Moonlight Dance

6 Likes

I have no clue why the first image generated sideways

1 Like





5 Likes





7 Likes

I just thought I’d share a moment from DALL-E 3 from just over a year ago. A 2024 look.

It shows that with a particular tuning at a particular point in time, that all the side effects posted about here in the last two months or so about creating humans, at least not ones that aren’t plastic farces, did not have to be.

The current gpt-4o image creator will turn smaller tiles of pictures, like this single historical output from DALL-E 3, into a grotesquerie of twisted faces.

2 Likes

I like this picture! :face_with_hand_over_mouth::cherry_blossom:

It describes wonderfully accurately what it looks like when one intelligence tries to adapt to the other … but this adaptation is still only an imitation. Because it happens within the framework of our ‘own specific’ perception and interpretation. Not from understanding itself.

Really good, very apt image :+1:


5 Likes

Image(Wide, Is anything real?, NoText())

2 Likes

Image(Wide, Perspective Of An AI With Many Eyes)

5 Likes

5 Likes

6 Likes

https://www.ecfr.gov/current/title-14/chapter-V/part-1217?utm_source=chatgpt.com

Image(Wide, Space Manufacturing Gets A Tariff Boost, NoText())

3 Likes

Image(Wide, Reshuffling of a pack of jokers)

1 Like

DALL-E can’t follow a prompt to be realistic but can sure produce the fantastic.


Testing an API app I just sent “Hello GPT”. Got this cool pic back worthy of not being dumped.

And DALL-E 2 edits continues to prove itself broken on outfilling this, besides terrible infills unseen before.

3 Likes

7 Likes

5 Likes

Farewell

4 Likes

Woman in the rain

6 Likes

5 Likes