(US & Canada) | AI Can Better Detect Cavities but Won’t Take Over a Dentist’s Role — Director of Advanced Analytics at Fortune 500 Retailer

Vivek Anand, Director of Advanced Analytics at a Fortune 500 Retailer, speaks with Derek Strauss, Chairman at Gavroshe and Editorial Board Member, CDO Magazine about the positive impacts of AI enablement, followed by the challenges and key takeaways for new AI and data science professionals.

Anand kicks off the third and final segment of the conversation by portraying the results of AI enablement across domains. First, he refers to a startup that is working on taking dental X-rays by integrating AI into the process.

Adding on, Anand believes that AI will succeed better in detecting certain cavities that do not meet the dentist's eye. However, AI will not take over the dentist’s role; instead, it will make the job easier.

Further, AI can carry out the redundant part of the job, leaving the more focus-oriented work for humans, says Anand. He asserts the need to keep humans in the loop for optimized outcomes.

Next, Anand shares an example of driving a Tesla and states how it conveniently predicts traffic and finds alternative routes, or during foggy situations. But it cannot be trusted completely when it comes to the intricate aspects, he adds.

From a challenge standpoint, Anand highlights the ethical and explainability aspects of AI. For instance, in the case of mortgage applications, a certain demographic gets rejected because the AI is taking input that does not represent that demographic population. Hence, the data is skewed.

At this point, explainability comes to the fore, and one must turn back to find the recommendation source and assess the biases in the model, notes Anand. Then, he discusses edge computing, which uses AI at the edge of an application.

Elaborating further, Anand says that if owners use a tool similar to Apple Vision Pro to dispatch fire trucks, the task fails because the tool removes the smoke. If the AI dispatching fire trucks is not trained on certain demographic data and geographical attributes associated with it, its performance will suffer.

In continuation, Anand mentions the data quality challenge, emphasizing how difficult it is to make the siloed data talk to each other. Last on the list, he stresses the talent shortage in this fast-evolving world of cutting-edge technologies.

The scale and adoption of AI depend primarily on the existing talent pool, says Anand, and to keep them relevant, talent improvement programs are needed.

Moving forward, Anand advises data science and AI newbies to leave behind the fear of missing out and focus on their business needs. While it has been established that data science and analytics can do numerous routine tasks better, understanding the business is key to that.

The first step is to run an exploratory data analysis to give the business notions a reality check. Furthermore, he reiterates the business need aspect, stating certain issues would require machines while others must involve humans.

For instance, in the case of product returns, machines can easily assess the situation and resolve issues, but a human is required to examine defective product cases. Therefore, it is critical to know what problem is to be solved and what it needs accordingly.

In conclusion, Anand urges building science that integrates across systems. He states that science is not a silo; it demands collaboration across stakeholders, and data science is a journey. As a last takeaway, Anand recommends starting small, iterating, and seeking feedback to accomplish objectives.

CDO Magazine appreciates Vivek Anand for sharing his insights with our global community.