Region 16 Has Big Plans for Digital Human “Emma”

Emma, Amarillo’s Digital Human, speaks over 60 languages and engages in real conversations.

Earlier this year, we did a spotlight on the Amarillo Public Library’s incredible digital literacy initiative. During that conversation, we learned about Emma, the city’s “digital human.” The first of its kind, Emma is an AI avatar, fluent in more than 60 languages, tasked with answering questions about city services and news.

Micheal Keough, the Chief Technology Officer of Region 16 Education Service Center, told us, “Rich Gagnon, the city assistant manager and chief information officer, is the individual that truly deserves the credit for Emma’s creation and vision as it stands today.”

We wanted to learn more about Emma, so we reached out to Michael. He is currently working on a partnership with the City of Amarillo to find pathways to use this project with schools in the future.  It turns out he has big plans to widen Emma’s scope into the community and area schools, making it of benefit to libraries as well.

LDO: Tell us about Emma.

Michael Keough: The Emma project was developed a year or two ago with the city of Amarillo and went to Dell. You can have Emma essentially communicate with anyone about the city in any language. It’s an expensive project; the cost is about half a million a year. For the education model, it costs about $175,000, which is still really expensive.

LDO: Talk about your plans to use Emma in education.

Michael Keough: We are taking it to stage two, using grants to purchase a server and house that in Region 16. Then libraries can use Emma in different capacities and use open source software to get Emma going.

My hope is that we can then make Emma a no-cost or low-cost resource for communities. By housing it locally for schools and other entities, we’ll be taking advantage of what the city is already doing. For schools that connect to the network, I can make sure all schools have access to a digital human like Emma, especially for those who don’t speak English. For students, Emma can also help them to do their homework. It all starts by getting a grant so we can purchase the server we need.

LDO: How does that work, once you get the grant?

Michael Keough: We have to find a path forward to start building and having the infrastructure to support Emma. The city owns the digital human model. They can give it to a non-profit. Region 16 would become a partner to bring Emma to others. Getting that in-house server is a huge part of it.

LDO: Besides the server, what are some of the other challenges with the project?

Michael Keough: Security is a big challenge. How old can kids be before they interact with AI? We have to be careful with AI hallucinations, too.

Of course, the fiber and the connectivity pieces are very important. E-Rate has been critical for that. Keep in mind, here in the panhandle of Texas, we have major connectivity issues. We have a GDP of 38 million, the same as all of Utah, but our infrastructure is similar to a third world country. We do have four resilient loops to go through all of the cities in the panhandle to keep things connected. It was important to build a traversable path, with education as a priority.

It’s going to cost $38 million for the fiber and we are working with local businesses, and hopefully the state, to come up with the additional $9 million it will take for the project.

LDO: How does Emma support digital literacy needs?

Micheal Keough: Language translation is a big part of it. It’s also great for the elderly who have trouble navigating their computers and find customer service difficult to deal with. A digital human can really help them. We did a pilot where Emma was used with some older folks. They would have entire conversations with it. It was a dynamic interaction. We want to pair that idea with a digital navigator program. We can pull somebody from the hispanic community or the elderly community to help them navigate.

We believe firmly in what we call “AI plus Human.” AI will never replace a human. There’s actually a dopamine response when you have a physical reaction with a human and you’ll never get that with a robot or AI, but Emma is close. Emma is fully audible and visual. Hit talk, and she talks back, and you have a conversation.

What’s unique with her is that she sways and moves, she makes expressions. She looks human when she talks. She’s very expressive.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.