Author(s)
Marco Iansiti, Luca Vendraminelli and Roberto Verganti
Source
Journal of Product Innovation Management, Vol. 37, No. 3, pp. 212-227, May, 2020
Summary
Increasingly, firms use artificial intelligence-based design (AI) to design interfaces and content displays. AI design is disruptive and changes the design process, but reinforces basic design principles.
Policy Relevance
Increasingly, human designers frame problems, and AI solves the problems.
Main Points
- Innovation requires decision-making to create ideas and solve problems, that is, "design;" until now, design was a slow, expensive human process, but AI uses algorithms and data to undertake design tasks.
- AI lets organizations overcome the limitations of human-intensive design by solving problems of scale and scope, and incorporating more learning and adaptation.
- Firms like Netflix and AirBnB automate problem solving, using AI to choose which interface or content to display to specific users.
- Hardware cannot be designed remotely in real time, but software can; Tesla embeds controls in digital interfaces and loads the cars with sensors to collect data for use by AI designers.
- The use of AI in design disrupts the process of design; what a user sees on screen is delivered and designed by a problem-solving loop powered by AI.
- Humans do not design solutions, but design the AI problem-solving loops.
- Designing, delivery, and use are simultaneous.
- The AI creates a solution for each user in almost zero time, at almost zero cost.
- When algorithms design, human designers focus on “sense making,” that is, in understanding which problems could be or should be addressed; this shift brings design closer to leadership.
- Surprisingly, AI design reinforces key principles of design.
- AI supports people-centered solutions, designed for every single person.
- AI supports the creation of many designs.
- AI design is iterative, as designs are continuously updated.
- Traditionally, the main limitation on people-centered design was scale, as it was unreasonable to design a solution for every use or every user; design was targeted to segments of the user base, or to the average user.
- Design supports human-centered development of AI; most tech startups think of their product as an app, but AirBnb thinks of the customer’s trip as the product.
- Some scholars equate problem solving and problem finding, and view framing a problem as a subset of problem solving; AI changes this perception, as problem solving is performed by machines, while humans frame the problem and drive the evolution of algorithms in a meaningful direction.
- Managers should reconsider how their organization innovates; the role of humans is not to develop full solutions but to understand what innovation problems are meaningful.