Alexandra Daisy Ginsberg / Machine Auguries

With AI-generated birdsong under an artificial dawn sky, Machine Auguries warns of our infatuation with technology at the expense of nature. Using thousands of field recordings of local and migratory birds, Alexandra Daisy Ginsberg has trained a pair of neural networks – a form of machine learning – to sing like these different bird species. We are no longer sure what is real as the chorus falls silent at the end of the work. Here at Bildmuseet we can experience three iterations of the chorus presented together for the first time: first dawn in London, 2019, then Toledo, Ohio, 2023, and from January 12 2025, a new site-specific Umeå version. The Umeå version will be installed on January 8-9, and the hall will be closed during these dates.
 
Alexandra Daisy Ginsberg (b. 1982) is a multidisciplinary artist based in London. Her work has been shown worldwide, including at MoMA, New York; the Museum of Contemporary Art, Tokyo; the Centre Pompidou, Paris; and the Royal Academy, London. This is the artist’s first solo presentation in Sweden.

Artworks in the Exhibition 

Machine Auguries, 2019 – ongoing
Multi-channel sound installation with programmed light, benches

Machine Auguries: London, 2019 (10 min)
Commissioned by Somerset House and A/D/O by MINI 
With additional support from Faculty

Machine Auguries: Toledo, 2023 (12 min)
Commissioned by Toledo Museum of Art and Superblue

Sarah Cook, Curator

We sense the changing of the seasons through sound and light. In Sweden, we hear the first cries of the curlew in spring and on dark winter mornings, we turn on our bedside lamp while we wait for the sun’s late rise. With its AI-generated dawn chorus of birdsong playing out in the light of an artificial dawn, artist Alexandra Daisy Ginsberg’s Machine Auguries warns of our infatuation with technology at the expense of nature.

The dawn chorus is a natural phenomenon that peaks in spring and early summer. Before and just after the sun rises, birds call to one another to find mates and defend their territory. In Ginsberg’s immersive installation, solo recordings of real birds are slowly taken over by synthetic birdsong  generated using machine learning, a form of artificial intelligence. This chorus is heard under an artificial sky that bathes us in the gloaming light of the predawn before taking us through sunrise and finally into the bright day. Following the architecture of the dawn chorus, the earliest birds are joined by later risers as the sun comes up, Ginsberg compresses this into a shorter composition than found in nature.

Machine Auguries is a site-specific series; each iteration simulates the experience of a particular place and moment in time. Working with local sound recordists, ornithologists, and sound archives, Ginsberg collates thousands of field recordings of bird species iconic to the location. These are used to train two neural networks working in a call-and-response process, a type of artificial intelligence called a Generative Adversarial Network, or GAN. Reflecting on how birds develop their song from listening to one another, each cycle or “epoch” of the machine’s learning is revealed as the GAN veers towards fidelity. As the dawn chorus progresses, the artificial birdsong becomes increasingly lifelike. By its conclusion, we can no longer be sure what is real and what has been lost.

Originally commissioned in 2019, Machine Auguries is a work of art that addresses the decimation of bird populations caused by human behaviours. The first iteration was created in London, where, as in many big cities, light and sound pollution are causing birds to sing earlier, longer, louder or at higher pitches to communicate. The second iteration was made in 2023 in Toledo, Ohio, a city located on the spring migration flyways from South America to the colder northern climes. Alongside pollution and habitat loss affecting bird populations, human-induced climate change is altering their migration patterns too. Only the species that can adapt will survive.

Observing the behaviour of birds to foretell omens, known as ornithomancy, has been practiced around the world for millennia. In classical Rome, augurs, or priests, would look at birds flying overhead and divine whether the gods approved or disapproved of the actions of humans. Machine Auguries offers a twenty-first century interpretation of this ancient practice.

We increasingly use AI to process digital data, but we must choose which information to process. Some researchers and technologists believe that AI could be used to conserve nature, for example to decipher animal communications to protect species. However, no technology is neutral, and their use depends on human actions. The same animal listening technology could be employed to exploit biodiversity further. The environmental crisis is a complex problem that requires societies to work together before technological intervention may be useful. Critics have noted that the growing use of AI depletes natural resources and consumes energy in ever larger server farms. This could negatively affect attempts to use it to protect nature.

The popular promises that humans can “save” nature through the development of new technologies are an unrealisable utopia as Ginsberg’s previous works illustrate. The Substitute (2019) reconstructs the functionally extinct northern white rhino, while Resurrecting the Sublime (2019) uses biotechnology to attempt to bring back the smell of extinct flowers. Other artworks have explored the human desire to exploit nature (The Wilding of March, 2019) or to optimise nature via technology (algorithmically-generating planting schemes for gardens designed for insects, not humans, in Pollinator Pathmaker, 2021 – ongoing). Ginsberg’s art does not promise that AI or emerging technologies are the panacea to deal with the difficult social and political work needed to address the environmental crisis. Her art does the opposite: it points out that we should be wary of relying on technology to help nature. It asks you to go outside and look after it. The Machine Auguries series celebrates the uniqueness of ecosystems but reminds us of their fragility and interconnectedness, evident when the same bird species reappear in different locales. By recording the impression of a place, Machine Auguries is also an expanding archive of vanishing local realities. As we experience the imitation of dawn inside the gallery, Ginsberg’s work reminds us of the impossibility of reproducing nature once it is lost. Bildmuseet has commissioned the latest version of the work – Machine Auguries: Umeå which will be added to the exhibition in January 2025.

This work is now part of the larger “Art and AI” research project at Umeå University supported by the Wallenberg Autonomous Systems and Software Program – Humanities and Society (WASP-HS). The Arctic context, and its climate, is of interest not only for reflecting on the changing migratory journey of different bird species found here, but also as it is now a site for the manufacture of batteries for electrical devices and for data server farms. The Art and AI project aims to show the value of artists and artworks in opening up conversations about the social and political implications of technological development. Machine Auguries brings into view one of the many paradoxes that AI presents to us: what are we asking technologies like AI to do for us and what role do we have in ensuring their sustainable use, for us and for the birds?

Curator Sarah Cook is UmArts WASP-HS Guest Professor at Bildmuseet and Umeå School of Architecture, supported by the Wallenberg AI, Autonomous Systems and Software Program – Humanities and Society (WASP-HS) funded by the Marianne and Marcus Wallenberg Foundation and the Marcus and Amalia Wallenberg Foundation. She is Professor of Museum and Information Studies at the University of Glasgow.

Birdspecies in Machine Auguries: London

Common chiffchaff

Great tit

Redstart

European robin

Thrush

Wren

Chaffinch

Dunnock

Wood pigeon

Willow warbler

Song thrush

Blackcap

Birdspecies in Machine Auguries: Toledo

American robin

Northern cardinal 

Indigo bunting

Gray catbird

Song Sparrow

Yellow warbler

Red-winged blackbird

Baltimore oriole

House wren

Warbling vireo

Black-capped chickadee

Common yellowthroat

Wood thrush

Tufted titmouse

Field sparrow

Prothonotary warbler

Mourning dove

Eastern phoebe

Ovenbird

Red-eyed vireo

Yellow-throated vireo

Swainson’s thrush

Blue-winged warbler

Rose-breasted grosbeak

Eastern towhee