Creating value with moderate AI

The first impulse on the subject of "Creating value with moderate AI" was provided by Prof. Dr. John Zimmerman from Carnegie Mellon University, who has over XNUMX years of experience in designing AI products and services. He shared his insights about the misconceptions and challenges of AI and discussed the critical involvement of designers and creatives in AI development. He began by noting that XNUMX% of AI initiatives fail to be deployed, with half of those that are deployed being unsuccessful for reasons such as unrealistic visions, lack of data, excessive cost, lack of user value or ethical concerns. Zimmerman suggested that a shift in strategy is needed, drawing on Google's approach to training teams to identify AI use cases. Rather than seeing AI as a superhuman entity and trying to make it replace experts, Zimmerman recommended targeting 'low-hanging fruits' to ensure that the tools being considered are feasible and provide a service that people want to use. The presentation highlighted that valuable AI models are often only moderately accurate, and advocated that design should act as a facilitator in AI, exploring areas where model performance can be improved.

Animacy and Agency

The second impulse presentation on “Animacy and Agency” was held by Dr. Madeline Gannon (atonaton), who specialises in the field of robotics, pushing the boundaries of what robots can do beyond their original design. Recognised as one of the top ten women in robotics industry by Analytics Insight, Gannon envisions a future where robots are curious, and alive, and kind. Her work emphasises animacy and agency, training robots to act more freely and improvisationally, and merging architecture and design with robotics. In her talk, Gannon challenged the traditional notion of robots, defining them not only as humanoid figures, but as any entity equipped with "a mechanical mind and mechanical muscles". She pointed out that robots in various forms, such as delivery drones, industrial robots and XNUMXD printers, are already integrated into our daily lives. Seeing them as mechanical creatures, she advocated a shift from controlling machines to connecting with them as they become more intelligent and autonomous. Her background in architecture is crucial to her approach, which aims to merge the arts and humanities with robotics.

Getting hands on with GenAI

In the third talk of the workshop, Savannah Kunovsky (IDEO) and Seth Trudeau (Routine Chaos) delved into the realm of Generative AI (GenAI), emphasising its role as augmented intelligence that can enhance human capabilities rather than acting as a rival. They underlined how AI has become seamlessly integrated into our lives, influencing both the design process and the potential for new products and services. Stressing the importance of designing responsibly, the duo explored how GenAI can open up new possibilities for designers, acting as a provocative partner that encourages bold and creative ideas. Kunovsky and Trudeau also highlighted IDEO's proactive research with Gen Z to understand their perceptions of AI. Their investigation showed that Gen Z see AI as a tool to deepen human connections rather than replace them, to help them navigate life with intention without shielding them from authentic experiences, and to assist them in embracing their imperfections rather than eliminating them. Finally, Kunovsky and Trudeau cautioned against bias, noting that while tools may seem morally neutral, the biases of their creators are reflected in their outcomes. They also warned of exploitation and monetisation, the dangers of misinformation and manipulation, and the importance of privacy, questioning the origins of the data used by models and how user input is managed. Their perspective therefore called for a careful and considerate approach to the integration of AI into our lives, ensuring that it serves to enhance rather than diminish human interaction and creativity.

In a final impulse, Yusuf Ahmad (Playlab) addressed the critical question of democratising AI creation, expressing concern over the current scenario where a limited number of individuals wield significant influence over AI tools' development. Highlighting that there are approximately XNUMX million educators globally, he argued for broadening access to AI tool creation to avoid a future dominated by products from a select few. Ahmad called for public infrastructure that empowers more people to develop AI tailored to their unique contexts, thereby increasing diversity and inclusivity in AI applications. Putting that into action, Playlab, the non-profit he co-founded, focuses on simplifying the process of creating AI tools and ensuring their accessibility and shareability. Ahmad emphasised the importance of technology enhancing life without adding unnecessary complexity or reducing quality. He advocated for accelerating open-source AI to foster transparency, noting that the opacity of most systems, which function as "black boxes," hinders understanding of their training and operational mechanisms. During the workshop, participants were invited to create their own application using Playlab and experienced first-hand the ease of building their own AI products through straightforward instructions and testing.

All impulses are also available as a short version on our YouTube channel.

The symposium

A symposium open to the public summarised the results of a learning-intensive day of workshops and invited the audience to participate in the discussion.

Each event of the series ends with a symposium, which is open to the public and takes place at the X-D-E-P-O-T of the Pinakothek der Moderne, where the findings of the workshops are shared and discussed with the audience. After a brief introduction by board members and hosts René Spitz, Annette Diefenthaler and Christoph Böninger, Prof. Dr. John Zimmerman recapped his talk from the previous day, sharing his thoughts on why so many AI initiatives fail and how designers and data scientists can work together to create tools that deliver valuable results. Subsequently, Dr Madeline Gannon outlined her vision of creating robots that are trained to go beyond simple, repetitive tasks to enable connection and mutual curiosity. This was followed by Savannah Kunovsky and Seth Trudeau's presentation, which focused on how Generative AI can enable designers to explore new possibilities, as well as some insights into how Gen Z perceive and want to interact with AI. Finally, Yusuf Ahmad demonstrated how to lower the bar for who can create with AI, and presented examples of applications created by workshop participants during his session.

An outlook on the Campus Project by iFDF x DIE NEUE SAMMLUNG until 2025: Upcoming events

The Director of Die Neue Sammlung, Prof. Dr. Angelika Nollert, concluded by thanking the panel and audience for the many thought-provoking ideas. There will be two more events in this series in XNUMX: in June (XNUMXth – XNUMXst), the topic will be "Leadership / Sustainability & Public Value", while the October event (XNUMXth – XNUMXth) will be dedicated to the subject of "Innovation".

If you are interested in participating in the invite-only workshops, please subscribe to our mailing list here (see below for newsletter registration). We will then invite you in due course. Participation is free of charge, the number of places is limited, and confirmations will be issued in the order in which registrations are received.