In order to create solutions that provide benefits to every demographic, brands are implementing initiatives to reduce inherent bias, racism or other discriminatory nodes

Inherent biases in programming have long plagued the tech industry. Though often unintentional, these biases can have the effect of excluding or turning off consumers. To reach more consumers, tech companies find it increasingly necessary to take it upon themselves to do what they can to create a more equal and inclusive world.

Accordingly, developers are adopting practical legislation and initiatives in order to remove unfair biases, racism or non-inclusive language barriers from artificial intelligence development and interfaces. Here’s how three companies are improving their artificial intelligence to make their services better suited to any and all consumers:

LG
The electronics manufacturer is working reduce inherent bias in its AI applications and aiming to create voice recognition software that can understand a variety of accents and dialects. For example, LG’s new voice recognition software can understand and respond to users from southern regions in South Korea, whose accents are often misunderstood by voice technology.

Google
Google is aiming to increase the amount of Google Assistant’s understood languages from 8 to 30 by the end of 2018, reaching over 85% of global Android users. Google Assistant will also support multilingual capabilities by the end of the year, allowing families who speak more than one language to interact seamlessly with the voice assistant.

Accenture
The consulting company created a professional resource and toolkit to assist in identifying and solving unfair bias within AI algorithms, before being deployed in large-scale AI solutions. The ‘AI Fairness Tool’ examines sensitive variables like race, age and gender within data models and measures if these variables are unfairly skewing the outcome of data sets. The tool also addresses false positives and false negatives in an algorithmic model and ensures that these are fairly distributed.

Focusing on developing inclusive technology is just one way that brands are improving their services and enabling an equal user experience. For more information on ethical improvements in the realm of data and AI, see PSFK’s report Implementing Ethical Standards for AI & User Data


Lead image: group of designers working together stock photo from Flamingo Images/Shutterstock

Inherent biases in programming have long plagued the tech industry. Though often unintentional, these biases can have the effect of excluding or turning off consumers. To reach more consumers, tech companies find it increasingly necessary to take it upon themselves to do what they can to create a more equal and inclusive world.

Accordingly, developers are adopting practical legislation and initiatives in order to remove unfair biases, racism or non-inclusive language barriers from artificial intelligence development and interfaces. Here’s how three companies are improving their artificial intelligence to make their services better suited to any and all consumers: