News for hackers. Stuff that matters.

wimpypistol

Re:Namecheap: Russia Service Termination on Monday February 28, 2022 @08:47PM (#30508164)
It really sucks that you'd do this.
Re:Keep Up: The first book written by an AI on Thursday August 13, 2020 @04:13PM (#24148139)
This is a word predictor fed on billions of tokens of text. Because it has been trained on human writing, it makes characteristically human mistakes (including spelling, grammar, and even some phonetic errors).
Re:Keep Up: The first book written by an AI on Wednesday August 12, 2020 @07:42PM (#24138417)
Incredible! I had no clue this existed. Thank you for pushing me down a rabbit hole for the last 45 minutes :-)
Re:Keep Up: The first book written by an AI on Wednesday August 12, 2020 @07:41PM (#24138412)
I was similarly surprised, but it's true. The original text is essentially the exact same (probably a 99.9% overlap) with the exception of a few spelling mistakes or grammar errors. My editing was about as much as you'd get from Microsoft Word, if that.
Hmm.. I may or may not speak French. Thanks for the idea.
Thanks for the input. I'm working on adding print functionality as we speak - let you know when it's up.
Absolutely! I was planning on providing this, but wanted to keep the model free of commercial services. There are open-source alternatives out there, though, which I may add support for in the future.
These are gorgeous! Great work.
Yes, good point. I think a major part of the appeal is the backstory. I do love abstract art, though, so I'm probably a non-representative sample :-)
Certainly a good point. Unfortunately, the output size is constrained by the training data (which in this case was ~800x800px). I think it would be kind of a pain to find 14,000 9000x9000 images, though, so I don't know if this will ever be able to fulfill this requirement. ;)
Thanks for the input! I'll add that functionality and let you know as soon as it's ready.
Definitely something I considered while feeding it a tremendous number of famous paintings (Rothko, Pollock, etc). Considering the nature of random seed generation/traversing through latent space, I don't see a strong legal case to be made for infringement. The law has done crazier things, though, so I'll keep an eye on it.
Beautiful.
Close! I used anime to test the basic functionality/UI of the site. Transfer learning was in fact done on the FFHQ512 dataset from Nvidia (I figured the complexity of its features was a little higher, which would lend itself well to abstract textures).
Completely fair assessment. Thanks for the input. I look up to Gwern quite heavily, so seeing myself referenced in the same sentence is nice - cheers.
This is correct. I agree with many of the commentators below - I could have been clearer. I simply didn't want to pay for a 24/7 GPU to generate on-the-spot (a little short on funds at the moment), instead opting to replace each series of 10,000 paintings once per month. Glad I could provide enjoyment to as many people as I did in the meanwhile!
Good thinking. Yes, that is correct.
Thank you. I love the clear, intelligent user interface of your site - it's given me a lot of inspiration for future projects. Keep in touch!
Correct. This is actually quite doable with word embedding or image classification of a topic, semantic analysis, and then providing those as input to a network like the one I used for 1SecondPainting. I love the future.
Excellent idea. I'll see what I can do.
This disturbed me in all the right ways.
Correct! I pre-generated 10,000 random seeds so I wouldn't have to rent a 24/7 GPU. The current plan is to upload a new slate of 10,000 each month.
I used plenty of Malevich while training, so that makes sense. Glad you enjoyed.
This was my reaction too the first few times I ran the network. I thought I was going crazy.
Beauty is in the eye of the beholder ;)
Haha yes, I first tested this with procedurally generated anime characters and must have forgot the first 10. Thanks for pointing it out.