Skip to content
Do you have adblock enabled?
 
If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(Ars Technica)   Stable Diffusion Image Synthesis was also the name of my grandpa's Progressive Rock band where he wired a Polaroid so he could control it with a Moog   (arstechnica.com) divider line
    More: Weird, Leonardo da Vinci, Mona Lisa, novel entertainment use of technology, deep learning image synthesis model, anonymous artist, fictional photorealistic character, Time travel, Stable Diffusion  
•       •       •

423 clicks; posted to STEM » on 21 Dec 2022 at 2:36 PM (13 weeks ago)   |   Favorite    |   share:  Share on Twitter share via Email Share on Facebook



4 Comments     (+0 »)
View Voting Results: Smartest and Funniest
 
2022-12-21 3:13:49 PM  
Far better than the Stable Diffusion porn I've seen so far.

Really, it's been disappointing.
 
2022-12-21 3:27:40 PM  
it amuses me greatly that people actually paid money to Lensa to create a model checkpoint that Lensa now presumably owns that they can use to create (currently fairly obvious) deepfakes of people doing whatever they want.
 
2022-12-21 4:52:08 PM  
Fark user imageView Full Size


Me walking my pet velociraptor. I played with a lot of prompts. "Napoleonic Admiral" is pretty dramatic looking.

replacementcool: it amuses me greatly that people actually paid money to Lensa to create a model checkpoint that Lensa now presumably owns that they can use to create (currently fairly obvious) deepfakes of people doing whatever they want.


Who really cares? Nobody is deep faking me. I can train models on my own PC anyway. Honestly, the amount of data to capture a trained model is pretty big, though you could extract the weights associated with the training target, but it's still a lot of data.

This is just a lot of fun. Kudos to that guy for having a great angle. Many years ago (mid 90s), I had fun photoshopping my son in various places, including floating in the space shuttle. AI (Machine Learning) just makes it easier and more interesting.
 
2022-12-21 5:58:13 PM  

LesserEvil: [Fark user image 512x640]

Me walking my pet velociraptor. I played with a lot of prompts. "Napoleonic Admiral" is pretty dramatic looking.

replacementcool: it amuses me greatly that people actually paid money to Lensa to create a model checkpoint that Lensa now presumably owns that they can use to create (currently fairly obvious) deepfakes of people doing whatever they want.

Who really cares? Nobody is deep faking me. I can train models on my own PC anyway. Honestly, the amount of data to capture a trained model is pretty big, though you could extract the weights associated with the training target, but it's still a lot of data.

This is just a lot of fun. Kudos to that guy for having a great angle. Many years ago (mid 90s), I had fun photoshopping my son in various places, including floating in the space shuttle. AI (Machine Learning) just makes it easier and more interesting.


I can make my own model checkpoint too. I sure as fark wouldn't pay to give it to some commercial interests.
 
Displayed 4 of 4 comments

View Voting Results: Smartest and Funniest

This thread is closed to new comments.

Continue Farking




On Twitter


  1. Links are submitted by members of the Fark community.

  2. When community members submit a link, they also write a custom headline for the story.

  3. Other Farkers comment on the links. This is the number of comments. Click here to read them.

  4. Click here to submit a link.