Demystifying AI: Experts Provide Insights on Rapidly Evolving Tech

The speed at which artificial intelligence is evolving and embedding itself into state organizations has necessitated a discussion establishing a basic understanding of the technology in government. Texas Department of Transportation (TxDOT) CIO Anh Selissen led such a discussion on the challenges and opportunities of AI implementation at the Texas Digital Government Summit* in Austin on Wednesday.

Joining Selissen were Google Cloud SLED Practice Lead for AI and Machine Learning Amina Al Sherif and Dell Technologies Senior Distinguished Engineer Travis North.

Regarding implementation, Al Sherif echoed sentiments they had previously shared during last year’s AI Day — organized by the Texas Legislature’s Innovation and Technology Caucus (IT Caucus) — regarding implementation and upskilling. According to Al Sherif, AI can be best implemented in back-office tasks, where there is less opportunity for public-facing system failures and more opportunity for human-AI task augmentation.

“Back-office tasks, tasks that typically would be taken care of by RPA, those are the tasks that we have found to be the most meaningful in terms of taking our workforce,” said Al Sherif. “Those are the tasks that we have found to be the most meaningful in terms of taking our workforce, and not only streamlining the workforce, but upskilling or forcing that workforce to upskill in some kind of way, shape or form to move on to the next cognitive thinking realm that you must be in.”

Selissen’s own agency has made a significant effort to streamline manual processes with AI, such as access management assignments and on- and off-boarding.

As far as upskilling is concerned, Al Sherif recommended agency leaders provide employees with the tools they need to learn and experiment with AI rather than struggling to keep up with the rapidly evolving technology.

“Make the tools available to them and just let them go,” said Al Sherif. “You have talented humans on your team. … Let them go out and consume information, but also let them know that if they’re trying to get ahead of the curve, nobody’s ahead of the curve.”

North cautiously advised attendees to only use AI in situations where alternate solutions are less effective to minimize risks and conserve resources.

“There were probably 800 different concepts of where we [Dell] were going to use it and we settled on eight,” said North. “And I will warn everybody, be very careful on how you use AI. If there is a closed form solution, it makes more sense to use your engineered closed form solution than to use AI to try to approach a closed form solution.”

The panel also addressed the known risks associated with AI regarding data privacy, especially when accessing AI tools through non-enterprise environments.

“If it’s free, you’re the product,” said Al Sherif. “When it comes to using tools, make sure that you’re paying for that privacy.”

Data management is key, North said.

“You have to be cognizant of the data you have because if ... you’re coming to a world where knowledge is manufactured, all of a sudden your business advantage is your data,” said North. “Having your data where it makes sense is really important and managing that is critical as well.”

*Note: The Texas Digital Government Summit is hosted by Government Technology. Government Technology and Industry Insider — Texas are both part of e.Republic.