Across all industries, business is now about technology and data. The sooner you understand and live that, the faster you will meet customer needs and expectations, create more business value and grow. It is increasingly important to reinvent business and use digital technologies to create new business processes, cultures, experiences and opportunities for customers.
One of the myths about digital transformation is that it’s about leveraging technology, but it doesn’t just mean that. To be successful, digital transformation inherently requires and relies on diversity. Artificial intelligence (AI) is the result of human intelligence, enabled by its talents and also susceptible to its limitations.
It is therefore imperative that organizations and teams make diversity a priority and think about it beyond the traditional sense. Diversity focuses on three key pillars:
People
They are the most important part of artificial intelligence; from the fact that they are the ones who create it. The diversity of people, the team of decision makers in the creation of AI algorithms, must reflect the diversity of the population at large.
This goes beyond ensuring opportunities for women in AI and technology roles. It also includes all dimensions of gender, race, ethnicity, skill set, experience, geography, education, perspectives, interests and more. Why? When you have diverse teams reviewing and analyzing data to make decisions, it mitigates the chances that their own individual and uniquely human experiences, privilege and limitations will blind them to the experiences of others.
Collectively, we have the opportunity to apply artificial intelligence and machine learning to drive the future and do good. That starts with diverse teams of people that reflect the full diversity and rich perspectives of our world.
Diversity of skills, perspectives, experiences and geographies played a key role in digital transformation. At Levi Strauss & Co, the growth strategy and artificial intelligence team doesn’t just include data and machine learning scientists and engineers. They recently reached out to employees across the organization worldwide and deliberately set out to train people with no prior coding or statistical experience.
They did not limit the background required; they simply sought individuals who were curious problem solvers, analytical by nature, and persistent in looking for diverse ways to approach business problems. The combination of existing expert retail skills and added machine learning knowledge meant that employees who graduated from the program now have significant new perspectives in addition to their business value. This one-of-a-kind initiative in the retail industry helps develop a diverse and talented bench of team members.
Data
Artificial intelligence and machine learning capabilities are only as good as the data that goes into the system. You often limit yourself to thinking of data in terms of structured tables (numbers and figures), but data is anything that can be digitized.
Digital images of jeans and T-shirts produced at Levi Strauss over the past 168 years are data. Customer service conversations (recorded only with permissions) are data. Heat maps of how people move through stores are data. Consumer reviews are data. Today, anything that can be digitized becomes data. There is a need to broaden the way we think about data and make sure that we constantly feed it all into AI work.
Most predictive models use past data to predict the future. It is often complex to have past data for reference. In fashion, you look to the future to predict trends and demand for completely new products, which have no sales history.
More data is used than ever before, for example, both images of new products and a database of products from past seasons. Computer vision algorithms are then applied to detect similarities between new and past fashion products, which helps predict their demand. These applications provide much more accurate estimates than experience or intuition, and complement previous practices with data-driven predictions and artificial intelligence.
Tools and techniques
In addition to people and data, diversity must be ensured in the tools and techniques used in the creation and production of algorithms. Some AI systems and products use classification techniques, which can perpetuate racial or gender bias.
For example, classification techniques assume that gender is binary and commonly assign people as “men” or “women” based on their physical appearance and stereotypical assumptions, meaning that all other forms of gender identity are erased. That’s a problem, and it behooves everyone in this space, in any company or industry, to prevent bias and advance techniques to capture all the nuances and ranges of people’s lives. For example, we can remove race from data to try to make an algorithm race-blind while continually guarding against bias.
The commitment to diversity in artificial intelligence products and systems use open source tools, being the most diverse because they are available to all and people from all backgrounds and fields work to improve and promote them, enriching them with their experiences and thus limiting bias.
The diversity of people, data, techniques and tools is helping Levi Strauss & Co. revolutionize its business and the entire industry, transforming the manual into automated, the analog into digital and the intuitive into predictive. It is also building on the legacy of the company’s social values, which stood for equality, democracy and inclusion for 168 years. Diversity in AI is one of the latest opportunities to continue this legacy and shape the future of fashion.