Digital Policy Salon: Online Shopping and Other Trends With Staying Power

Mairead Matthews and Rosina Hamoni

AI development teams generally comprise domain experts, business experts and AI/systems experts. Each of these players brings unique core skills to the table. Domain expertise is critical. In financial services, this could be retail banking, insurance or credit card fraud, while in health care, this might mean diagnostic imaging or automated treatment planning. A business expert’s core skillset, on the other hand, is knowing how to find the best product-market fit. Finally, the AI/systems expert has the core technical skills needed to develop a successful AI model.

However, ICTC researchers also found that siloed skillsets on an AI development team is not enough. Individual team members (or departments in larger organizations) also need to know enough about the other two knowledge areas to successfully communicate and collaborate. Without this overlapping expertise, AI development teams risk creating products that are not well-suited to the realities of their industry or do not provide the best product-market fit. In worst case scenarios, business and domain experts are unable to communicate to technical developers what products they want or need, while technical experts are unable to articulate what can and cannot be achieved using AI.

Critically, these challenges aren’t unique to AI. Across past projects focused on virtual and augmented reality, and current projects focused on agriculture technology and digital health, ICTC found that both business and domain knowledge are crucial for these software development teams. The following key informant quotes related to digital health expand on this requirement:

Read the full article here 📝

On June 29th, 11 am-12 pm EDT, ICTC in partnership with DIACC invites you to join Namir Anani, ICTC President and CEO, for a virtual panel engaging with North American experts and learn more about digital IDs and their role in an increasingly data-driven economy.

Register here 🗓

The technology needed to fulfill orders is costly for stores, and the workers who pick items off the shelves often feel the pressure of being tracked.

Nick Fraser would regularly start his grocery picking shifts at the local Fred Meyer store at 4 or 5 a.m.

Upon arriving at the Kroger-owned chain, Mr. Fraser, who lives in Salem, Ore., would begin fulfilling online orders, zipping through aisles with a hand-held device and scanning bar codes on everything from cereal and milk to chicken and vegetables. The screen on the device was his guiding light. His goals: to retrieve each item within 30 seconds and to find 95 percent of a customer’s grocery list.

Loading the virtual shopping cart has never been so easy, and people are doing it more than ever—especially for their groceries. Online grocery shopping has increased by an estimated 700% since the pandemic began. Among the many changes brought on by the pandemic is more widespread use of online tools to facilitate grocery shopping. Initially, this was because people wanted to maintain social distancing, and grocery stores remained among a handful of critical out-of-home activities. But now, as restrictions lift and post-pandemic life is around the corner, many shoppers are sticking with the “click and collect” option.

The article highlights how the rapid proliferation of online grocery shopping has outpaced retailers’ ability to manage it, and this is already significantly impacting the responsibilities of grocery store employees. While increased demand may imply a need for new hires at grocery stores, estimates suggest that “a typical grocery store could already operate with up to 55%-65% fewer staff hours.” Whether technology will make these roles easier or less common (or both) is a key yet unanswered question. Automated checkouts appear to only have been the beginning of what is bound to be a proliferation of automation in grocery stores. For more on Intelligent Retail, see the ICTC blog post From Innovation to Imperative. - Khiran O'Neill