Peter Read Peter Loved is an artificially intelligent artwork made by Ali, AImade.art’s newest iteration of Artificial Intelligence.
Presented is a Halloween themed painting made by A.I. The painting was more of an inspired collaboration of some fan photos of pumpkins and fall festivities. The desired effects were achieved by pairing the photos with the machine learning capable with Ali and we have to say, some interesting and spooky results emerged!
To create it, Ali used algorithms from neural style transfer libraries on Github as well as some sandboxed code that interprets text based on preset values called VQGAN+CLIP.
Using thousands of paintings from renowned artists and publicly available images, Ali used the techniques for vector quantized Generative Adversarial Network (GAN) and Contrastive Language for image pre-training. GANs are systems where two neural networks, (1) a generator which synthesizes images into different styles, and (2) a discriminator which tries to determine if the image is a valid creation. The system feeds back on itself to incrementally improve the piece.
The result is this one-of-a-kind artwork carefully selected and minted as a non-fungible token (NFT) in Ethereum's blockchain to certify its uniqueness and ownership. The owner receives an unlimited license to use this artwork, including for commercial purposes.
With the purchase of this NFT the first buyer will receive a physical copy of this artwork printed on museum-quality canvas and shipped worldwide at no extra cost.
In order to offset the carbon emissions this transaction generates, We have partnered with Sea-Trees to remove over 1 ton of CO2! For every NFT sold we will:
- Sequester 1/2 ton of CO2 with carbon credits from the Southern Cardamom REDD+ Project, Cambodia which provides 200+ jobs, education and healthcare benefits for 16,000+ people in the local community, and produces 3.5 million metric tons of carbon credits per year.
- Plant 2 mangrove trees in Indonesia, with Eden Reforestation Projects which has the potential to sequester an additional 1/2 metric tons of CO2
- Restore 1/2 sq-ft of kelp in Palos Verdes, California. This project provides habitat and food for over 700 species of algae, invertebrates, and fish. It also reminds us of My Octopus Teacher
Physical copy specifications:
• Size: 12x12'' (30.48x30.48 cm). 1.5″ (3.81 cm) deep
• Acid-free, PH-neutral, poly-cotton base
• 20.5 mil (0.5 mm) thick canvas
• Canvas fabric weight: 13.9 oz/yd² (470 g/m²)
• Printed on textured and fade-resistant canvas (OBA-Free)
• Mounting brackets included
• Hand-glued solid wood stretcher bars
• Algorithms inspired by VQGAN+CLIP and Style Transfer libraries
• Trained with over thousands of public domain paintings from publicly available/royalty free images on the Internet, the Met Museum and WikiArt as well as publicly available/royalty free images on the Internet