When Robbie Barrat prepare an AI to canvas and reproduce classical nude paintings , he expected something at least recognisable . What the AI produced instead was unfamiliar and unsettling , but still intriguing . The “ paintings ” see like material body - like ice cream , spilling into pool that only mistily recall a woman ’s dead body . Barrat severalise Gizmodo these meaty blobs , distressing and unplanned as they are , may touch both artistic production and AI .
“ Before , you would be give the computer a set of rules it would execute absolutely , with no way for interpretation by the computer , ” Barrat said via email . “ Now with AI , it ’s all about the machine ’s interpreting of the dataset you feed it — in this event how it ( funnily ) interpret the bare portrayal I eat it . ”
AI ’s influence is certainly more pronounced in this undertaking than in most computer generated fine art , but while that was n’t what Barrat intended , he says the results were much expert this way .
“ Would I want the results to be more realistic ? Absolutely not , ” he sound out . “ I need to get AI to generate unexampled case of graphics we have n’t seen before ; not force some human position on it . ”
Barrat explained the process of discipline the AI to develop imagery of a curving consistence from some surreal parallel universe :
“ I used a dataset of thousands of nude portraits I scraped , along with techniques from a new composition that recently came out called ‘ reformist Growing of GANs ’ to generate the images , ” he said . “ The generator tries to bring forth painting that fool the discriminator , and the discriminator tries to determine how to tell the conflict between ‘ bogus ’ picture that the generator feeds it , and real paintings from the dataset of naked portraits . ”
TheFrancis Bacon - esquepaintings were purely serendipitous .
“ What bump with the naked portrait is that the generator figured it could just feed the discriminator blobs of shape , and the discriminator was n’t able to tell the divergence between strange blobs of flesh and humankind , so since the source could systematically put one over the differentiator by painting these strange variety of flesh instead of naturalistic naked portraits ; both components stopped learn and get better at picture . ”
As Barrat pointed out on Twitter , this method acting of work with a reckoner program has some art history common law . induce an AI do the artist ’s specific direction is reminiscent ofinstructional art — a conceptual arttechnique , best exampled by Sol LeWitt , where artists bring home the bacon specific instructions for others to create the artwork . ( For example : Sol LeWitt’sWall Drawing , Boston Museum:“On a paries surface , any continuous stretchiness of paries , using a hard pencil , place fifty stage at random . The points should be evenly distributed over the area of the wall . All of the points should be connected by straightforward lines . ” )
If you re conversant with Sol LeWitt- I feel like AI produce nontextual matter has a relation to his study . He would demonstrate a set of rules , and then someone else would render those rule and make a drawing from that . I feel like I m laying down rules that the AI interprets and bend into artistry .
— Robbie Barrat ( @videodrome)March 28 , 2018
Giving the AI limited autonomy to create art may be more than just a fallal , it may finally go to a really Modern form of generating art with exclusively fresh subjectivities .
“ I desire to use AI to make its own new and original nontextual matter , not just get AI to mimic things that people were take a crap in the 1600 ’s . ”
DesignDigital nontextual matter
Daily Newsletter
Get the good tech , science , and culture news program in your inbox daily .
News from the future , delivered to your present tense .
You May Also Like
