this post was submitted on 08 May 2024
1131 points (98.7% liked)

Science Memes

10348 readers
1477 users here now

Welcome to c/science_memes @ Mander.xyz!

A place for majestic STEMLORD peacocking, as well as memes about the realities of working in a lab.

Rules

  1. Don't throw mud. Behave like an intellectual and remember the human.
  2. Keep it rooted (on topic).
  3. No spam.
  4. Infographics welcome, get schooled.

Research Committee

Other Mander Communities

Science and Research

Biology and Life Sciences

Physical Sciences

Humanities and Social Sciences

Practical and Applied Sciences

Memes

Miscellaneous

founded 2 years ago
MODERATORS
[email protected]
[email protected]
[email protected]
1131
Old comic, more relevant than ever (www.smbc-comics.com)
submitted 4 months ago by [email protected] to c/[email protected]
74 comments fedilink hide all child comments
 

https://www.smbc-comics.com/comic/ai-9

you are viewing a single comment's thread
view the rest of the comments
[โ€“] SomeAmateur 94 points 4 months ago* (last edited 4 months ago) (5 children)

I had to run that through Bing AI real quick lol

[โ€“] brbposting 61 points 4 months ago (1 children)

Ataliative ๐Ÿ˜ฎ

Adetvi Learning ๐Ÿ˜ฒ

Blowv the ciompetittio ๐Ÿคฏ

[โ€“] [email protected] 25 points 4 months ago* (last edited 4 months ago) (1 children)

BlocklBerach

Clould frist lustion

Agรญee

[โ€“] brbposting 14 points 4 months ago (1 children)

Clould FRIST

[โ€“] [email protected] 5 points 4 months ago* (last edited 4 months ago)

Edit: Fixed

[โ€“] [email protected] 36 points 4 months ago (4 children)

Iโ€™m surprised itโ€™s able to make readable text.

[โ€“] zalgotext 22 points 4 months ago (1 children)

Blovw the competittio

[โ€“] [email protected] 8 points 4 months ago

Blovw it hard!

[โ€“] [email protected] 8 points 4 months ago

Image generation models are generally more than capable of doing that they're just not trained to do it.

That is, just doing a bit of hand-holding and showing SDXL appropriately tagged images and you get quite sensible results. Under normal circumstances it just simply doesn't get to associate any input tokens with the text in the pixels because people rarely if ever describe, verbatim, what's written in an image. "Hooters" is an exception, hard to find a model on Civitai that can't spell it.

[โ€“] ghen 7 points 4 months ago

The worst part is that it now tries to add text to a whole lot of pictures

[โ€“] SomeAmateur 4 points 4 months ago (1 children)

Yeah it's been improving over the past few months. It's hit or miss though

[โ€“] [email protected] 5 points 4 months ago

I like to post sometimes on the "guess the song" AI communities, but more often than not, the Bing image creator just plast the lyrics on the image making it useless for the game.

[โ€“] [email protected] 23 points 4 months ago

Yeah these look exactly like things I'd see on billboards in Vegas when certain conventions are in town...

[โ€“] [email protected] 11 points 4 months ago (1 children)

I will probably use these images in a corporate PowerPoint. I'm not asking for your permission, I'm warning you. Sorry, it's too good. (I will credit you as a CTO of some company ending in -SYS or - LEA if you want)

[โ€“] SomeAmateur 6 points 4 months ago

Lol go for it just send me a screenshot of the slide!

[โ€“] Grandwolf319 6 points 4 months ago

Ahh I miss the days when people who knew and used generative AI understood they are best for shits and giggles.