Co-Chairs：Sang Suh, Radmila Juric

Design thinking has been around since 2008 but did not become the main development practice in software development and design. The specificity of software development, underpinned with computer science, with constantly changeable software and communication technologies, requires models of software abstractions and architectures in order to deploy them with suitable software technologies. Thus, it would require revisiting and more thinking on how exactly Empathize, Define, Ideate, Prototype and Test form design thinking would accommodate current practices of software development.
Data science has exhibited a sharp rise, which started less than a decade ago. However, it has not materialized as a discipline with its principles, roles, and paradigms. The fast rise of data science has become its curse. Apart from having numerous, elaborative, powerful algorithms which can process an enormous amount of data in no time, we do not find anything else in data science. We create predictive inference of variable accuracy, sometimes with biased models and results, and often unexplainable in terms of their computational values. We do not wish to undermine the data science ability to process an excessive amount of data and get “some meaning from it”, but we should be aware that predictive inference has limitation’s and, as statisticians used to say “could become dangerous instruments in wrong hands”. If we add to this the fact that predictive inference, with learning technologies, define what is known today as Artificial Intelligence, then the current practices of creating software solution based on AI must be investigated further. Could design thinking help in finding out where we stand with AI? Could it eliminate data science and predictive inference shortcomings.
In this workshop we look at the relationship between design thinking and data science form different views and debate possibilities of either merging or complementing them. It is difficult to predict where data science will take us in near future, because we started seeing resistance towards underpinning business decision on predictions only and would prefer help form logic inference. On the other side, how do we promote design thinking in this dynamic field of computing where everything is changeable, moveable, wireless, automated, human augmented with machine and vice versa (to mention just a few characteristics of modern computing).

Our aim is to revisit the design thinking process for intelligent systems development with AI, machine learning, and data science. Past experience and achievement among the panelists will be shared and synthesized to anticipate or suggest future directions to support intelligent systems design through sustainable and trustworthy AI.

Panelist will lead an open-discussion using presentation slides.