This is the seventh article in our series, The Power of Technology.

Generative artificial intelligence admittedly drives a lot of conversations about the trust that people should or should not put in AI. But the government can easily make the case for AI and machine learning’s responsible use, said Tony Celeste of Ingram Micro.

Why? Because the government has been an early adopter of AI technologies, said Celeste, executive director and general manager for Ingram Micro Public Sector. By way of example, he pointed to the NASA Mars rovers, which used AI in numerous ways. “Those systems were in development many, many years before something was launched and then had to travel all the way there,” he said.

The government should “educate citizens about how they’ve been using this technology for decades, and really spend some time talking about narrow AI uses — how it’s cataloging and categorizing information and bucketing, making it more efficient and saving tax dollars,” Celeste said. “Generative AI is really just an expansion of that.”

Issues of trust arise because “people don’t understand AI,” said Art Villanueva, chief AI/ML architect for Dell Federal during a discussion with Celeste for the Federal News Network The Power of Technology series.

Both agreed that transparency is critical. Villanueva pointed to the need for “explainablity” when using AI, a term that developed during the earliest uses of machine learning and Natural Language Processing. Explainability “basically means, ‘How did your machine learning algorithm or model get to the solution that you’re asking?’ ” he said. “If people don’t understand how that works, then you have trust issues.”

Celeste and Villanueva each shared opportunities that they see for broader AI/ML use within government in the near future.

Opportunity 1: Document management

“Using AI for document management, there’s a ton of possibilities,” Villanueva said. He identified three: classification, ordering and summarization

AI/ML can be used to speed access to the right information by sorting documents — classifying them — based on an ML model. “When you have a gazillion documents that you need to pour through, where do you start if you need to learn something?” he said. “Which documents do you do start with?”

There’s also document ordering. “If you want to learn a particular branch of knowledge, which documents do you want? Which documents do you do read first? And which ones do you do read later? Because one document may or may not be a prerequisite for another one,” Villanueva said.

Then, there’s summarization. That comes into play if someone identifies a vast store of documents and decides they need information from all or many of them, he said. “There’s just no way for a human to get all that information, to absorb all of it.” A summary can help someone grasp the key elements and point to additional reading or research that may be needed.

Opportunity 2: Institutional knowledge

Celeste raised the idea of using AI/ML to capture the institutional knowledge of federal employees and also for training new crops of employees.

“Coming out of the pandemic, we saw lots of change in workforce,” he said, with people leaving who weren’t expected to leave and taking with them large swaths of institutional knowledge amassed over 20-, 30-, 40-, even 50-year careers in some cases.

“Mapping out an organization — its people, its processes, its functions, its roles — and building on that data set to allow for knowledge management is a way to preserve that information that individuals gain over time,” Celeste said. Plus, that store of institutional knowledge can then be mined using AI to help new workers get up to speed far more rapidly and without disrupting services to the public.

Technology meets vision when it comes AI/ML

The future of AI is bright in government, both Villanueva and Celeste said.

Even now, an apt question is, “What does AI and ML not touch?” Villanueva said. “Because everything, everything within government, is going to be affected by AI and ML.”

Agencies will continue to successfully scale the use of AI/ML, Celeste predicted, lauding initial efforts to foster trust and develop governance. In particular, he noted the Office of Science and Technology Policy’s work on an AI Bill of Rights blueprint that addresses ethics and bias issues of the technology’s use.

“We have the processing, storage and networking capacity, and thus the data repositories, to actually start really putting the power of AI and machine learning to work,” Celeste said, adding that it will make providing speedy and valuable insights possible. “It’s going to improve our interactions and our life experience and overall well-being.”

To read more articles in The Power of Technology series, click here.

Learn more about Dell Federal’s offerings through Ingram Micro Public Sector on Xvantage.

Listen to the full show:

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.