What are you looking for?

Experimentieren mit ChatGPT & Co

19.06. – 23.06.2023

Experimenting with ChatGPT & Co.

Seminar type
Thematic seminar, for schools, for trainees
Target audience
Ages 16 and up
focus
Digitalization and Democracy
Language
German
Everyone's talking about artificial intelligence, ChatGPT, and the like. But what's really behind it all, and how do we want to deal with AI? What do training data for artificial intelligence have to do with lived societal diversity? Can algorithms discriminate? Does social media promote racism, or does it create spaces to combat it in solidarity? What skills and knowledge do we need in a digitized society to be able to act responsibly and with awareness of prejudice in everyday life and at work? – These are many complex questions that we need to negotiate as a society. Only by grappling with them can digital civic participation become possible.

In June, trainees from the Louise-Schroeder-School explored these and other aspects of digitalization – in the seminar
„Lifehack Digitalization – #howto Living Diversity in the Digitalized World“.
To achieve this, they researched, experimented with text- and image-generating AI, discussed the topics, and developed their own products – both analog and digital. Results from the four workshop groups offer insights into their work process.
Analog Black and white photography, Podcast, Documentary photography/poster campaign and analog game development.

Podcast/Audio Workshop

What exactly is ChatGPT? What dangers are associated with AI? How has it changed? The history of artificial intelligence up to the development of ChatGPT has been completed, and what does the future of AI look like? The self-developed and produced [product/service] provides research results, opinions and insights into AI experiments.
Podcast „MeMo: Wannsee AI“ – published on Audiyou.de, enjoy listening: Link on Audiyou.de to the podcast „MeMo: Wannseen AI“

(Please note that by clicking on the following images and links, data will be transmitted to Flickr.com data will be transmitted; see the link to the privacy policies of the services and Privacy policy of the wannseeFORUM.)
For their podcast title image/cover, the participants generated images using AI, which can be seen in a video on Flickr released album.

Used Statistics:

Statistics on the use of AI in everyday life in Germany, 2020: https://de.statista.com/statistik/daten/studie/1176437/umfrage/nutzung-von-kuenstlicher-intelligenz-in-deutschland-im-alltag/

Further sources:

  • Federal Agency for Civic Education (n.d.): Timeline of Artificial Intelligence: A Journey through the history of machine thinking, bpb.de, [online] https://www.bpb.de/fsd/h5p/zeitstrahl-ki/ [Accessed on June 20, 2023].

Workshop group
Analog black and white photography

Photography from the photo series „World Collector“

Our understanding of the world and the formation of opinions are strongly influenced by the perception of images – companies like Open.ai, for example, have developed bots with their Midjourney service that generate images based on textual specifications. Do algorithms thus change how we perceive and interpret visual content and our relationship to images as a form of documentation? Do they reproduce and reinforce prejudices? What images could encourage us to consciously reflect on our handling of prejudices and to invite others to engage with them, instead of assigning value to people and groups, and to see diversity as a celebration of differences and commonalities?
In the workshop, participants use their own analog image production to explore questions about the construction of reality through photographs by artificial and human intelligence.

They developed photographic series and presented them in an exhibition they conceived and implemented themselves.
Parts of the photographed exhibition can be seen here (pdf).

Analog Game Development

What role plays Artificial intelligence for self-determination and democratic values in everyday life, the professional world, and social interaction? The participants of this workshop playfully explored different perspectives on the opportunities, fears, and dangers of AI: What tasks does AI take on? How is it trained so that human errors—such as lying or prejudice—and anti-democratic values are not reproduced and reinforced? Using text- and especially image-generating AI, they developed and produced their own games, which could be tested during the final presentation. Some of the game materials are published here:

„Flabopulies“Lie or truth? Decide! (PDF game materials)

„Neural Nexus“ This will lead you into the exciting world of artificial intelligence!
You all share the same goal: to become the best AI. Information and data are exchanged, stolen, and negotiated. But beware – a human is trying to harm you. Your mission: Feed the server with information
(pdf game instructions)tung)

Documentary photography/poster campaign

Assignment for the task „Poster campaign“

A few days before the seminar began, the European Parliament debated the Artificial Intelligence (AI) Act, the most comprehensive regulation of artificial intelligence worldwide to date. The extent to which AI pursues capitalist power and profit interests, or public-interest-oriented societal goals, and the way it will ultimately act depends heavily on the future legal framework for its creators. The initially "voluntary code of conduct" must also address issues related to the promotion and reproduction of structural discrimination. This includes the visual language of the media. How are racism, prejudice, and bias disseminated or combated when image-producing AI draws from existing digital sources? What linguistic stances in prompts generate which images? How can one's own position on AI in a diversity-oriented society be shaped through photography (and perhaps even co-determined by publishing it as a source of imagery for AI)?

The workshop participants experiment with their own visual concepts as well as with AI services.

Photo by Hannah Lawen and Marie Nguyen for the Prompt
„Something You Can’t Live Without“

In the exercise „Promptroulette“, an AI randomly generated terms, to which participants then photographed their image ideas – with an exciting comparison to the results of an image-generating AI.

For documentary photo series, they explored Berlin's urban space to see where and how digitization/AI already defines/changes spaces.

Photograph from the series by Piet Sykora and Tom Eichberg

Documentary series

In theImpressions from the seminar on Flickr

The seminar was supported by the Senate Department for Education, Youth and Family.
It took place as part of the collaboration in the program „Civic youth education in the AdB“ in the project group „Social Question and Civic Participation“ instead, funded by the Federal Ministry for Family Affairs, Senior Citizens, Women and Youth.
The wannseeFORUM Foundation is a recognized sponsor of the Federal Agency for Civic Education.

Cover image: created with AI by the podcast workshop group