DAIN Studios Germany CEO Dirk Hofmann and Senior AI Consultant James Keirstead spoke last week at IXDS’s Pre-Work Talk in Berlin and Munich on the theme “Opportunities and Pitfalls of Prototyping with Artificial Intelligence”. IXDS is a leading innovation and service design studio in Germany, dedicated to ‘prototype the future’. Their monthly Pre-Work Talks bring together a community of service designers, engineers and researchers to learn about the latest developments in the field.
During and after the presentations the audience asked insightful questions about how to build organisational capabilities that support AI prototyping, how to handle setbacks in AI prototyping, and the role of humans in AI-augmented systems. It seems there is a demand for further elaboration on these topics, so stay tuned for our upcoming posts!
For now, here are the highlights of the events:
Today advances in AI technologies and platforms make it easier than ever to develop products and services that respond intelligently to users and their contexts. Aside from all the hype and buzz from deploying AI across various markets, industries and realms of business, one might still wonder when does it actually make sense to use AI and is it even the right solution for a given problem? As a start, the following questions can be very useful for assessing whether the current state of AI could be useful in prototyping:
- Describe; how would a “theoretical” human perform the task today?
- If the human expert was to perform this task, how would you respond so that they could improve the next time they were to perform the task?
- If a human were to perform this task, what assumptions would the user want them to make?
The different use cases for prototyping with AI were explored together with data sources needed for those – be it new data or designing with existing data. As data is the foundation of AI, in order to leverage the opportunity of prototyping with AI it requires a new, more intensive model of collaboration between service designers, data experts and strategist, not to forget the end users – from the very beginning.
The approach for AI use in design, or Data Driven Design as Dirk put it, has close parallels with the Design Thinking approach applied by IXDS.
However, while the barrier might be low to get started with Data Driven Design and the available tools, one should still be aware of the most common pitfalls:
- The Hype: Start right away with the AI solution and skip over the meaningful problem to solve.
- Believe in the Black Box: Lack of understanding about licensed algorithm results limits the potential to innovate;
- Lack of Data: a model that was not trained on the right data with the right goal, leads to lack of stability and usability of results;
- Perfect World Driven: Created solution fails to resolve the most critical edge cases, and therefore fails to build trust and engagement;
- Fail to Scale: lacking the leadership teams, infrastructure and the adaptation of relevant processes to continuously evolve;
- Try to Shortcut: Trying to leapfrog the process of data collection, algorithm development, and testing.