Perverse Affordances (2018)
負擔不良 (2018)

About The Artwork
About The Artwork

Perverse Affordances (2018)
負擔不良 (2018)

Sarah Friend (Canada)
莎拉.芬地 (加拿大)

A generative adversarial neural net is a model in machine learning that can generate new images based on a dataset. Perverse Affordances begin with one was trained on 10000 screenshots of global social media platforms and asked to dream new screenshots, and also encompasses an interactive online sketch based on these images. In some ways, it is a new social media network made in collaboration with this algorithm.

The processes by which we identify and imagine images happens on a level of our awareness that is, at least most of the time, invisible and pre-conscious. In contract to this familiar biological process, artificial intelligence is profoundly alien – visible. Watching the images created by generative adversarial neural nets evolve gives a rare opportunity to see “seeing” itself.

This piece turns the alien eye of machine learning on the interfaces that mediate our online lives. Like seeing itself, these interfaces are ubiquitous enough to have become invisible, and when using them we perceive them only as proxy for one another. But of course, this belies their constructedness. Interfaces are not transparent. They cannot give us pure proxies to other humans. What about a network allows us to confidently say it is social?

Much like a perverse incentive, which is emergent behavior within a system that contradicts the intentions of its designers, Perverse affordances invites us to consider interfaces as systems that have been designed – and the assumptions they contain or conversely do not contain, about the humanity of their users.

生成對抗神經網絡是一個可以根據數據集而衍生新的圖像的機械學習模型。作品《負擔不良》開始於在一萬個全球社交媒體平台的屏幕截圖上「學習」並生產出新的截圖,並包括根據這些圖像而同時生產的互動在線草圖。在某方面來說,這是一個與數字運算合作而生產的新社交媒體網絡。

我們人類在識別和想像圖像的過程通常發生在我們的意識之上,是不可見的,也是不可預料的;而在這類同的生物過程中,人工智能是完全的異種,一切都是可見的。觀看以生成對抗神經網絡所創的圖像之時,提供了一個難能可貴的機會去看見「看見」本身。

這件作品把機械學習中的奇異點轉化成我們線上生活的介面;就像「看見」的本身,這些介面無處不在,無容忽視,而當我們使用它們時我們只會把它們視為彼此的「代理」(proxy)。但這個理所當然地掩蓋了它的構造,而介面是不可能透明的。它們不可能提供單純的「代理」(proxy)給其他人,那麼一個網絡如何容許我們可以自信地稱呼它為「社交」的?

就像一個不正當的誘因,即是在一個系統裡與設計者反動的新興行為;作品讓觀眾去思考把介面變成已設計完好系統的可能性,以及它們包含或相反地不被包含用家人性化的假設。

Sponsored by 贊助

Artist Biography
Artist Biography

Sarah Friend (Canada)
莎拉.芬地 (加拿大)

Sarah Friend (@isthisanart_) is an artist and software engineer working at a large blockchain development studio. When not doing that, she creates games and other interactive experiences. Her practice investigates murky dichotomies - like those between privacy and transparency, centralization and decentralization, and the environment and technology - with playfulness and absurdist humour. She is a proud Recurse Center alum, and has presented at Transmediale in Berlin, Ethereal Summit in NYC, and NorthSec in Montreal. She was recently chosen as one of Canada's 30 under 30 developers, is one of the organizers of Our Networks, a conference on all aspects of the distributed web in Toronto.

莎拉 ‧ 芬地 (@isthisanart_)是一名藝術家,同時亦是軟件工程師,工作於一家大型區塊鏈發展工作室。工餘之時,她會創作遊戲以及各種互動體驗。她的創作實踐用有趣與荒謬式的幽默去探究模糊的二元論,比如說在隱私與透明之間、中心化與去中心化之間,還有環境與科技等。她是Recurse Center的榮譽校友,作品曾參與柏林的Transmediale、紐約的Etheareal Summit和加拿大蒙特利爾的NorthSec。她最近被選為加拿大30位30歲以下的軟件開發者之一,是Our Networks – 一個多倫多有關分佈式網絡的國際會議。