Arti­fi­cial intel­li­gence is occupy­ing soci­ety. This is no news, but can be seen in the most diverse areas of every­day life. For example, through the grow­ing num­ber of exhib­i­tions about and with AI. For our mini-series AI in Museums, we vis­ited two exhib­i­tions in the imme­di­ate vicinity.

Our series star­ted at the Kun­st­mu­seum Stut­tgart. There, the exhib­i­tion "SHIFT. KI and a Future Com­munity" was on view from 4 Feb­ru­ary to 21 May 2023.

Idea, Aim and Back­ground of the Exhibition

The exhib­i­tion is a joint pro­ject of the Kun­st­mu­seum Stut­tgart and the Museum Marta Her­ford. It is ded­ic­ated to the dia­logue between art and sci­ence in the field of arti­fi­cial intel­li­gence. The exhib­i­tion was cur­ated by Friederike Fast, Dr Eva-Mar­ina Froitzheim and Ann Kristin Kreisel.

Behind the pro­ject is the obser­va­tion that AI, as a key tech­no­logy of digit­al change, has a decis­ive influ­ence on the social, polit­ic­al and eco­nom­ic con­texts of soci­ety in the present and the future. This change, this shift — as the exhib­i­tions name already dis­plays — is to be shown and ques­tioned with­in it. In addi­tion to the influ­ences on our every­day lives, AI also always raises the ques­tion of the under­ly­ing images of human beings and con­cepts inter­twined with them, such as intel­li­gence, con­scious­ness and creativity.

The exhib­its focus on the inter­lock­ing of AI and the real­ity of life, the pos­sib­il­it­ies, lim­its and eth­ic­al respons­ib­il­it­ies that arise from this.  They thus open up a per­spect­ive space for vis­it­ors to exper­i­ence and think through these questions.

The exhib­i­tion was accom­pan­ied by an extens­ive edu­ca­tion­al pro­gramme with sym­po­sia, guided tours and lec­tures. It was developed in cooper­a­tion with the Stut­tgart Centre for Sim­u­la­tion Tech­no­logy and Cyber Val­ley Stuttgart/Tübingen.

know­botique in cooper­a­tion with Pablo Alar­con and Ira Wil­helm, Nina Bandi, Pablo Torres, Anna Brom­ley, Joerg Fran­zbeck­er, Leila Pea­cock, Sally Schoen­feldt, Ian Wooldridge: AMAZONIAN FLESH — how to hang in trees dur­ing strike?, 2018–2019

Jenna Sutela: Indigo, Orange and Plum Mat­ter (I Magma), 2021 

Heath­er Dewey-Hag­borg and Chelsea E. Man­ning: Prob­ably Chelsea, 2017 

Struc­ture

The exhib­i­tion is made up of the indi­vidu­al works of eight artists and artists' col­lect­ives, each of whom has developed their own scen­ari­os in which AI becomes vis­ible and can be exper­i­enced by the senses. The indi­vidu­al exhib­its are loc­ated in dif­fer­ent rooms that extend over two floors of the Kun­st­mu­seum Stuttgart.

The arrange­ment of the indi­vidu­al works does not fol­low any per­cept­ible nar­rat­ive strand or prin­ciple of order, but rather cla­ri­fies the basic idea of the exhib­i­tion, which is to make indi­vidu­al aspects of the changes in soci­ety that are advan­cing through AI tan­gible and to stim­u­late crit­ic­al debate.

The Per­spect­ive on AI

Vis­it­ors exper­i­ence the dimen­sions of AI through the mod­el situ­ations that the eight artists and col­lect­ives cre­ate in their works. Each indi­vidu­al mod­el situ­ation ques­tions the lim­its and pos­sib­il­it­ies of AI and the eth­ic­al respons­ib­il­ity that arises when deal­ing with it.  The works each focus on spe­cif­ic aspects and applic­a­tions of AI. These include deep­fakes, large lan­guage mod­els, data streams and oth­er vari­ants of machine learning.

The group of works "Repräsent­an­tinnen (2021)" by Louisa Clem­ent, for example, asks about the effects and lim­its of human-machine rela­tions. For this, the artist has giv­en sex dolls her own appear­ance and voice, fed them per­son­al data, chat his­tor­ies and answers to intim­ate ques­tions. Vis­it­ors can inter­act with one of these dolls. Through the con­ver­sa­tions, the fig­ure learns and expands its lin­guist­ic abilities.

In their anim­ated film "in vivo — in vitro — in silico (2023)", the artist duo kennedy+swan focus on a com­pletely dif­fer­ent aspect of AI. The pos­sible inter­weav­ing of AI with bio­lo­gic­al intel­li­gence in the form of xeno­bots derived from the cells of clawed frogs. Their film shows a dysto­pi­an image of the future in which these xeno­bots in the year 2043 scurry through a human body as "buddy bots" and take over tasks such as detox­i­fic­a­tion etc. from the inside, which are con­trolled from the out­side by an app. An applic­a­tion that is sup­posed to offer massive health bene­fits but would only be access­ible to rich people.

These two examples illus­trate what char­ac­ter­izes the entire exhib­i­tion: a very strong focus on humans and a con­stant anthro­po­morph­iz­ing of AI. There is little or no pen­et­ra­tion into the real back­ground and actu­al char­ac­ter­ist­ics of AI. Rather, the artists seem to take up aspects and applic­a­tions of the AI dis­course in their works that are asso­ci­ated with strong fears and neg­at­ive ideas for humans. These aspects are an import­ant part of the dis­course around AI but could be bet­ter clas­si­fied and thus more adequately dis­cussed by explain­ing the tech­nic­al real­ity. How­ever, the exhib­its seem to miss this oppor­tun­ity time and again.

Sci­ence Com­mu­nic­a­tion Perspective

The basic idea of the exhib­i­tion, to dis­cuss the changes tak­ing place in soci­ety through AI on the most diverse levels and to open up spaces of exper­i­ence, holds great pos­sib­il­it­ies from a sci­ence com­mu­nic­a­tion per­spect­ive. Oth­er forms of com­mu­nic­a­tion, such as art, can con­trib­ute to pro­mot­ing vis­it­ors' own engage­ment with the top­ics and thus enable them to enter into the dis­course on AI. In addi­tion, groups of people who are oth­er­wise less con­cerned with tech­no­lo­gies also gain access to this topic.

To do this, how­ever, it is neces­sary to explain the basic con­cepts of this dis­course and to present the back­ground and cur­rent res­ults of research on AI in a com­pre­hens­ible way. How­ever, this aspect is neg­lected in the SHIFT exhib­i­tion. There is a gloss­ary on indi­vidu­al basic terms of AI. How­ever, this is only vis­ible after scan­ning a QR code, which rep­res­ents an addi­tion­al bar­ri­er, espe­cially for those people who are not very tech-savvy. The indi­vidu­al exhib­its use AI tech­no­lo­gies and applic­a­tions that are not explained fur­ther and often have only remotely to do with AI. This is com­poun­ded by the often anthro­po­morph­ized depic­tion of AI in the indi­vidu­al works.

The poten­tial to open up a more in-depth view of arti­fi­cial intel­li­gence and to con­trib­ute new impulses to the dis­course remains unused. AI con­tin­ues to be per­ceived as the inscrut­able, indefin­able "black box" that is some­how related to humans and could some­how pose a threat. Clas­sic images and nar­rat­ives, such as glass heads in which lumin­ous mat­ter can be seen or humanoid robots read as female, as shown in the exhib­i­tion, fur­ther under­line this missed opportunity.