How to Make Sure AI Devices Pass Ethical Muster
In approaching this problem of social responsibility, software designers have begun to develop their systems through two ethical lenses.
First, they have adopted the approach used by medical ethicists who involve humans in studies associated with new clinical trials and medical procedures. These safeguards have been in continuous development since the Nuremberg Code of 1947 and include informed consent by the human subject and transparency in the potential adverse reaction an individual may have to the trial involved. They give special consideration to the involvement of minors and vulnerable members of society who may not themselves be able to make informed decisions. These considerations are directly translatable to the development of AI in our homes. Strong corporate social responsibility informs the customer that the device is always observing us; asks for “informed” consent; provides consumers accurate knowledge and ownership regarding data retention, while also protecting minors.
The second lens is the use of design thinking, where the product development passes through five stages: Empathy, Define, Ideate, Prototype and Test. The first of these, the empathy stage, is crucial in aligning AI products with social dynamics. In this stage, designers aim at gaining a deep understanding of the actual user experience with the product. This differs from traditional design processes as AI solutions are algorithmic, based on data that may have elevated levels of bias from inappropriate sourcing or application.
Without such design considerations, society will be increasingly uncomfortable with AI, and turn away from it, sadly failing to reap the significant benefits it can provide.
See the full story here: https://blogs.wsj.com/experts/2019/06/28/how-to-make-sure-ai-devices-pass-ethical-muster/