mindmelt.party
— “real-time hallucinatory music visualizer and networked culture explorer” from Nullsleep and friends via @nullsleep #Vanity Fair interviews Billie Eilish for the sixth year
— this will be her last annual video for a while, she plans to keep recording them but stagger the releases via MetaFilter #Movetodon, find your Twitter friends on Mastodon
— new tool with a streamlined UI; I’m [email protected], if you want to join me #Why Japanese web design looks different from the rest of the world
— don’t miss Sabrina’s accompanying article and resources #The Barnacle Goose Experiment
— Everest Pipkin’s latest is an “abiogenesis body horror idle clicker” via b3ta #Marine Beaufils’ videogame needlepoint
— especially love the series inspired by The Sentinel via Web Curios #Stable Diffusion 2.0 released
— trained from their new OpenCLIP encoder, with higher resolution, a 4x upscaler, and a new depth-guided img2img model #Defunctland’s documentary on who wrote the Disney Channel theme
— Kevin Perjurer solves a mystery with impeccable production values and a meditation on legacy #Meta made an AI that can win at Diplomacy
— building extremely persuasive AIs that are more effective negotiators than humans, what could go wrong #Taylor Lorenz interviews @Dril
— a rare interview, showing his willingness to commit to the bit until the cleansing fire burns it all down #Can AI generate the perfect Thanksgiving?
— Priya Krishna makes GPT-3 generated recipes and it goes about as well as you’d expect via @[email protected] #ooh! directory
— Phil Gyford made a wonderful categorized directory of 852 (and growing!) active blogs #Interactive Fiction Competition 2022 winners announced
— 70% of the top 20 are choice-based instead of parser-based, continuing the trend of the last few years via @IFTF #Corridor Crew on neural radiance fields and the future of 3D scanning for video
— NeRFs are neural networks that can generate complex 3D scenes from 2D photos via Garbage Day #Tim Molloy’s production stills for TV shows that never existed
— profoundly weird generations made with Midjourney v4 #Ask My Book
— Sahil Lavingia made an experiment where you talk to a vocalized AI of him about his book #The Verge’s deep dive into the murky waters of copyright and fair use in AI
— I talked to the reporter for this piece, which does a good job covering some of the complex questions around legality #Bureau of Multiversal Arbitration
— multiplayer Discord game tasking you with finding objects in the multiverse with an AI image generator #Bubbles
— charming bubble-blowing simulator, the rest of Saharan’s portfolio is great too via Web Curios #Mastodon’s Eternal September begins
— thoughtful post on the different cultural norms and expectations between Twitter and Mastodon #Tumblr launches Important Blue Internet Checkmarks
— a real steal at two for $7.99, with up to 24 per user #Inside the Twitter meltdown
— Platformer reports on Elon Musk’s ongoing self-inflicted train wreck #Abortion Bus
— fundraiser livestream of a Desert Bus-esque simulation of the 23-hour drive South Texans need to make to get an abortion #The Infinite Conversation
— an AI-generated never-ending discussion between Werner Herzog and Slavoj Žižek; more details on Hacker News #Josh Millard finishes his grandfather’s stained glass project
— I followed every step of this as he documented it on Twitter since learning about it in 2019 #AI-generated art sparks furious backlash from Japan’s anime community
— very relevant to my new post #Diana Smith’s pure HTML/CSS Varga-style pinup
— inspect the source for some comedy, and be sure to try it in older browsers #Invasive Diffusion: How one unwilling illustrator found herself turned into an AI model
Last weekend, Hollie Mengert woke up to an email pointing her to a Reddit thread, the first of several messages from friends and fans, informing the Los Angeles-based illustrator and character designer that she was now an AI model.
The day before, a Redditor named MysteryInc152 posted on the Stable Diffusion subreddit, “2D illustration Styles are scarce on Stable Diffusion, so I created a DreamBooth model inspired by Hollie Mengert’s work.”
Using 32 of her illustrations, MysteryInc152 fine-tuned Stable Diffusion to recreate Hollie Mengert’s style. He then released the checkpoint under an open license for anyone to use. The model uses her name as the identifier for prompts: “illustration of a princess in the forest, holliemengert artstyle,” for example.
