"Aw yeah it's time for cookies," writes AI ringmaster Janelle Shane (previously at BB).

One neural network I use, called textgenrnn, tries its best to imitate any kind of text you give it. I’ve given them paint colors, band names, and even guinea pig names and in each case their results are somewhat… mixed. (Paint colors called Stanky Bean, Stargoon, and Turdly, for example) The problem is that it doesn’t know what any of these words mean - it’s just picking letter combinations that seem likely to it.

This is what happened when I gave it all the cookies from a list of American recipes. This is what human cookies sound like to a neural network.

Previously, previously. Read the rest

Dan Hon (perfecting earlier work by Tom Taylor) trained an AI on the vast corpus of British place names, then set it loose. The results are amazing, illustrative of an uncannily human humor seemingly at work, something you'd never get from the standard syllable-randomizing place name generators of yore.

"There aren’t as many cocks as you’d think," he writes.

My favorites: Brotters Common, Boll of Binclestead, Farton Green Pear End, Weston Parpenham, Sutsy Compton, Stoke of Inch, Kinindworthorpe Marmile, Rastan-on-croan, Fuckley, Fapton, Waterwitherwell.

See also Hon's AI trained to generate Ask Metafilter post titles.

Surely neural-net-generated Liffs are next. Read the rest

Research scientist Janelle Shane writes: "I’ve been training a neural network (based on this open-source neural network framework from Andrej Karpathy) on datasets from recipes, to lists of Pokemon, to superhero names. I decided to see if it could invent names for new fruit varieties - I fed it a list of apple, peach, pear, plum, and cherry varieties, and asked it to generate more."

Sunbrown Stanker Pork Gala Horspest Shivercell Hencough Moregall Brown Soften Ruby Wally Ruck Nagtort Blee Red Redcells Zuby Glong Zeelcher Hacker Gala Soften Fuji Klunk 134 Horking Read the rest

Google's neural net is amazingly good at figuring out what you draw. In this game, it correctly guessed five out of six doodles I drew: cookie, saw, scissors, beach, grass. It missed watermelon. Read the rest

Take one Google Inceptionism neural-net system, which, when fed its own output over and over, begins to hallucinate dogish-lizardoids in random noise; add one supercut of cheese-porn pizza ads; stir thoroughly and strain. (via JWZ) Read the rest

Since Cory posted about the Deep Dream image recognition algorithm last month (and Rob earlier today) it's inspired an explosion of iterations like Roelof Pieters' DeepDreamed Fear & Loathing in Las Vegas. Read the rest

In "Inceptionism," scientists at Google Research describe their work training neural nets with sets of images, then tweaking the "layers" of neural net nodes to produce weird outcomes. Read the rest