Commentary: AI Will Optimize (Not Replace) the Work of Government

The AI rush is slowing down a bit now and in the near term, and in the future we will begin to see the actual value added of using AI tools. For now, what we can note is the value added to enhance routine business workflows and provide better information to the public.

For decades — some might argue centuries — the globe has been tackling day-to-day problems and coming up with solutions to workflows’ basic constraints. Those efforts are not lost with AI. All those tools and systems built to solve problems in our daily lives will become augmented by AI technology, though they all won’t be entirely replaced. Call it optimization.

For example, AI technology was recently added to the Placer County “Bob the Bot” chatbot solution on the county webpage. It does not replace the existing system and strategy; rather, it augments it. This new version allows for better contextual results and responses, leveraging the entire conversation flow, instead of one-liners from web pages, which it offered previously. This results in using AI to provide a more human-like conversational experience.

To top that, the new system can also understand 50-plus languages while also converting responses to match the language in which the questions are asked. Today, the most successful solutions that include AI are using the technology as an augmentation to an existing system/platform — an upgrade rather than a replacement.

Again, we started without AI, and with no Internet, and put a lot of time and discovery into how things should be processed. AI augmentation brings us better solutions to problems we’ve already found ways to solve. Converting legacy systems to inherit new augmented AI solutions that can leverage language translations, humanized conversation and emotional intelligence will enhance expectations and yield better results.

For example, presentation of an AI augmentation can be seen in many chatbot solutions today. It is important to note that AI only augments these solutions. There are many attributing factors that play major parts in a well-rounded solution that can be enhanced by an AI augmentation. These factors include, but are not limited to, good content and good context — and, when speaking of context, leveraging the full conversation stream. Chatbot solutions have been around for years now; however, augmenting them with AI produces better human-like conversion flows, translation and communications in various users’ native languages, regardless of the native language used by the jurisdiction.

AI today uses a very contextual base, meaning enough information needs to be provided so that the solution can respond properly. This, however, leads to the problem of hallucinations. This is a problem we will always be battling when using context-driven AI platforms. Now that we see how AI will best augment systems, the real work begins!

I have to agree with others that AI policy cannot come before understanding how AI works. The Government AI Coalition of San Jose aims to attack that very problem. While the coalition focuses on providing tangible resource templates for various agencies, the goal is not to provide documentation that answers the problems for everyone, but to provide a good starting point. To provide the best results, the coalition has members including industry leaders from all levels and organizations nationwide in what I would call a never-before-seen collaboration effort, which is extremely difficult to accomplish in government!

The Coalition is open to working with vendors to engage and understand better ways of incorporating policy and well-thought AI acquisition/procurement processes. Further, there will be an upcoming summit on Dec. 4 in San Jose. Vendors may want to contact the coalition regarding sponsorship; see the contact details on the event announcement.*

*The Summit is being hosted in partnership with Government Technology and the Center for Public Sector AI. Both are part of e.Republic, Industry Insider — California’s parent company.