by ckozlowski on 6/6/24, 12:54 PM with 71 comments
by thegrim33 on 6/6/24, 3:58 PM
Whether there's even any data behind this or if it's an article based on a handful of random tweets who knows, as it's a pay-walled article.
On a meta level, what percentage of HN readers are paying for a subscription to the washington post? Why are links to paid articles even allowed when the vast majority of readers won't even have access to read it?
by thot_experiment on 6/7/24, 12:44 AM
Even if we have strong, even overbearing legislation put into place to protect artists, we're just going to end up with the biggest most profitable offenders buying/using illegally trained models through middle men, and feigning ignorance when/if they are found out.
Though most of the "has this model been trained on X" conundrum is likely to be irrelevant on a years (months?) timescale as art styles are far less unique than artists would like to think. See tencent's PhotoMaker or other modified CLIP approaches. Even a model not trained on a particular face can be conditioned to generate said face using a oneshot approach because the vector representing that face can be constructed even if the face itself doesn't exist in the training data. I'm certain the same is true for artistic styles.
by exabrial on 6/6/24, 5:02 PM
by astrange on 6/7/24, 12:09 AM
by paxys on 6/6/24, 10:38 PM
by shrimp_emoji on 6/6/24, 7:41 PM
by qarl on 6/7/24, 1:05 AM
https://www.theverge.com/2012/12/20/3790560/instagram-new-te...
by iamacyborg on 6/6/24, 9:46 PM
by jerrymcflurry on 6/7/24, 10:32 AM