The Future of Art: Deep Dreams, AI and ascribe
A few weeks ago, the Gray Area Foundation for the Arts and Research at Google hosted a benefit auction and exhibition in San Francisco titled DeepDream: The art of neural networks. The exhibition was a unique one. One made up exclusively of artworks created using Google’s artificial neural networks.
If you’re wondering what exactly “artificial neural networks” means, Google wrote a blog post about how this all works and it’s nothing short of fascinating. To summarize it briefly, what happens is that different layers of the technology are set up to spot patterns or shapes, everything from corners and edges to trees, animals and eyes. Whenever the network spots one of these elements, it interprets and enhances it. In the post, they make a comparison to watching clouds and picking out familiar shapes. They call this “Inceptionism.”
The Deep Dream site is worth the visit. Head to http://deepdreamgenerator.com, upload an image and marvel for yourself at the strange beauty that it spits back at you. Greg McMullen, our CPO, put a photo of me into the generator, the result of which he has rather appropriately named “Puppy Medusa.”
In the same way that the sound of yesteryear’s dial-up Internet connection could be thought of as digital files trying to speak through a line meant for audio, these images have been described as visual representations of what the dreams of machines look like.
The entire limited edition collection of works was auctioned off with the proceeds — totalling nearly $100,000 — going to the Gray Area Foundation for the Arts, which works to bridge the worlds of art and technology.
At ascribe, we embody this intersection of art and tech ourselves, which makes our collaboration with Gray Area on this project all the more exciting.
ascribe is a service that allows creators — artists, photographers, writers, graphic designers, musicians — to register, transfer and track their digital content. Our goal is to give control back to the creators. We do so by providing the tools to lock in authorship so that work is always properly attributed, create limited digital editions and set specific licensing parameters for seamless consignments, loans and transfers of work to galleries, museums or marketplaces. We also provide visibility into where work spreads online by showing when and where work has appeared. For more on that, check out WhereOnThe.Net. Essentially, we provide the closest possible approximation to ownership on the Internet. By registering work with ascribe at the time of creation, you can set your work free into the wild World Wide Web with ease and confidence. In addition to the amazing artists that we work with on a daily basis, we collaborate with galleries, online marketplaces, collectors, museums, archives and prizes. And now, Gray Area.
All of the Deep Dream work at the exhibition was registered with ascribe. In this case, the work was uploaded to ascribe and one unique digital edition was created for each piece. The master files that were registered had the associated legals, licensing and the artists’ information attached to it, including full name, name of the piece, year created and so on. They were then consigned to Gray Area for the exhibition so that they could be sold on the artists’ behalf. Technical information was also attached to the registered artworks when consigned so they knew exactly the kind of material, framing and dimensions to use.
Those who made purchases at the auction now have a Certificate of Authenticity (COA) for their piece with a built-in cryptographic ID. These COAs offer a clear provenance, verified authenticity and can be printed out.
At ascribe, the future of digital art and digital collecting is an exciting one, especially with the advancements in areas like artificial intelligence and virtual reality. Gray Area is the latest in a long list of collaborators who have chosen to use ascribe’s technology and tools to ensure that the artists they work with can properly protect and manage their artwork. Sometimes, it’s just that easy.
