In the AI of the Beholder

Eric Heling, MBA
Product Architect Team Leader, Garmin International

Education: Michigan Technological University; University of Wisconsin; University of Missouri-Kansas City

AI Mission: Helping people understand technology is here to assist us all. And making sure it does just that.

Eric Heling works in Area 51. No, we're not talking a highly classified government outpost or a secret alien base. Instead, Heling's Area 51 is a special projects group within Garmin's Olathe, Kan., headquarters. The multinational technology company specializes in GPS, but Heling dabbles outside that focus. He uses his background in electrical engineering and signal processing to zero in on possibilities for using technology to build the next Garmin business or product that will enrich customers' lives. Heling sees this kind of think-tank, incubator approach as a way to help companies like Garmin evolve with the changing times – and help customers. His team strives to launch products that solve customers' problems and make life easier.

Heling's philosophy about technology: If you're passionate, you'll find a way to make new ideas work. Here's his advice on making technology work for you.

His perspective on AI is based on that customer-centric focus. Here it is in his words.

— Bowman Report

Heling's words: The future is now

Most people don't realize how much artificial intelligence (AI) is already in their lives. You can't avoid AI. It's on your phone, in your car, on your doorbell and in your television. You can't outsmart it or avoid it. You must learn how to use it to your advantage. Most people are probably already doing this without even knowing it. As a product designer, I strive to create a seamless technological experience that doesn't disrupt your life. The product shouldn't encumber you or slow you down.

Take image recognition, which is the AI I'm most excited about right now. Consider why facial recognition – a form of image recognition – is included on a phone. You can use it to log onto your phone, but the uses go beyond that. The external camera could include scene recognition capabilities that could judge your mood to evaluate mental health or survey the contents of your refrigerator and suggest what to cook.

Know the maker

Problems with AI arise when variables are present. In cases with few variables, the data is organized into buckets of easily identifiable information. But a case with variables outside the normal buckets can quickly throw off a system. The best AI or machine learning systems can handle these edge cases or variables.

To create quality systems, engineers need to train the computer algorithm using a large variety of samples. For example, when we create AI systems that will be used for humans, we need different body types, ethnicities, ages and so on. Creating a standard model of a male in decent health is easy. It's when you throw in variables that AI gets complicated.

I've tricked myself a few times when testing an AI system. I get excited when it works on me. Then we use it for a case involving a child or a senior citizen or a female, and the system breaks down.

That's why the only negative I see with AI could be the person who programmed the machine. When you choose AI, you need to choose companies you trust. Most established companies are offering AI, like IBM, Google, Amazon, Microsoft, Apple and, of course, Garmin. You could say these company names add credibility. All credible companies are trying to associate their qualifications with their AI technology.

When it comes to privacy and platforms collecting information and data about you, you need to be smart. Be cognizant of the companies you allow into your life and into your home. Be smart about the information you provide and share.

See the good

For example, a lot of video games are collecting player information. Some people are concerned about this. On the flip side, my kids and I view it as an advantage.

If the game learns us, it can make the playing experience more enjoyable. Hence more addicting, right? But you don't necessarily need to play more. If the games are learning you, they can present new things you might like. They also can adjust the difficulty level to meet your skills instead of raising the levels until the player is frustrated and quits.

I believe we can use games with AI technology to teach children. A child needs to be encouraged but not overwhelmed, an AI game can do that. Again, AI is about improving someone's life. It's exciting to think AI could improve the way children learn and increase their enrichment and enjoyment by adapting to their interests.

AI also can adapt to help people set limits, and I recommend people incorporate those features into their AI use. Smart phones include the do-not-disturb feature to help prevent disruptions by learning your sleep patterns. Your phone may soon be able to sort your messages by learning what's important to you. If it's important, like a call from family, then the phone will let it in even during the middle of the night.

Change for the better

I think all problems, from the mundane to those that might overwhelm society, can be solved via tech and progressive change. In the animal health industry, I think AI would help manage workflow with a shortage of veterinarians. A simple AI chatbot could help by being available 24/7 — which people want and often expect — to triage simpler cases. This would allow veterinarians to work more regular hours and focus on solving truly complex cases.

This is better than a pet owner looking for answers by firing up a search engine, which is AI too by the way. What if instead that pet owner could get quality, curated information from a real expert or, better yet, a machine trained by many experts?* The pet owner could type a question into an intelligent veterinary platform anytime day or night. The platform will ask for information from the pet owner then provide a response. What if a quick picture or video of the animal could be used? Pet owners would think their veterinarian is awesome because it's as if she's accessible all the time. The pet owner would still feel the strong relationship to their vet based on the expert and immediate service they experience.

As a technology guy and a pet owner, I think all veterinarians should embrace AI regardless of whether they're new in the profession or seasoned doctors. I think everyone can – and will – embrace AI slowly, evolving and adapting little by little. Veterinarians possess a lot of knowledge and with that, they probably can compete with AI to some extent. But as processing power and cloud storage get cheaper, they won't be able to compete forever.

*What if has become what is

Platforms already exist to provide pet owners with always-on veterinary advice. Bowman Report contributor Cal Lai leads Ask.Vet, which powers MyPetDoc, an online subscription service that allows pet owners to live chat with a licensed veterinarian.

Leave a comment