If enacted, the legislation proposed by Michigan Democrat Gary Peters and Indiana Republican Mike Braun would be the first law addressing the training of the federal workforce since Congress passed the Government Employees Training Act in 1958. The renewed attention to capacity-building at the federal level should spark interest among state and local government leaders and start a conversation about what public professionals need to know about artificial intelligence, especially widely accessible generative AI tools like ChatGPT.
With Microsoft’s announcement that it will make generative artificial intelligence available to government customers, there is an urgent need to go beyond asking only how to regulate the technology and to ask how public professionals can responsibly use AI in their work. Healthy skepticism about the risks is invaluable, but there is not enough attention yet to addressing how we might use AI to support improved government functions and services.
We are already seeing the potential for what AI can do. Because generative AI can make it faster and easier to create text, images, videos and even music, for example, we can use it to help write policy memos, as the city of Boston recommends in its guidance to public servants.
The technology is also helpful for rapidly turning legalese and government-speak into plain English. Innovate Public Schools, a California-based nonprofit dedicated to parent advocacy, is working on such a project, exploring using generative AI to make the language in a family’s individualized education plan easier to understand, especially for low-literacy and non-native English speakers. Imagine taking advantage of the power of generative AI to break down complex language into simpler terms, automatically translate text into multiple languages, and extract and summarize important information. With 15 percent of public school students across the country studying under an IEP, the impact for families could be transformative.
That gets at a larger problem, one that AI tools are well-suited to address: From filing for unemployment to applying for nutrition assistance, accessing government services is often hopelessly complicated. It’s all too easy to miss out on benefits to which one is entitled because of a paperwork error. In New Jersey, where I serve as the state’s first chief innovation officer, we have already seen the benefit of simplifying forms. Despite an explosion of unemployment claims during the pandemic, by changing language on the application, among other refinements, we were able to reduce the time to apply by 48 minutes.
Using AI-powered chatbots, government agencies can also make simplified and much-needed government information available to residents 24/7. For example, during the pandemic in New Jersey, we partnered with the Federation of American Scientists to crowdsource answers to the public’s COVID-19 questions from a network of more than 600 scientists. We then uploaded the information onto Amazon Alexa so that the public could ask questions by voice and get answers at their convenience.
Most exciting of all is the move by governments worldwide to experiment with new AI-enabled participation platforms to foster greater public engagement. Digital democracy is getting a boost from AI. Iceland’s Citizens Foundation, which created the Better Reykjavik platform on which over half the city’s population has offered ideas for how to fix urban problems, now incorporates an AI assistant that helps city leaders and the public summarize what their fellow citizens are saying on the site. Meanwhile, Romania’s ION searches social media to understand what that country’s citizens are talking about and care about and reports back to the country’s leaders on this sentiment analysis.
But to accelerate the many unexplored opportunities for using generative AI to improve governance in this country, governments at all levels need to invest in training people in the technology. Unless public professionals know what AI is and how to use it, we will miss out on the chance to use it for public benefit.
The good news is that building public-sector capacity doesn’t have to come at a significant cost. In New Jersey, we launched a free training program for public servants in innovative skills, including new technologies, data analytical thinking and human-centered design. We use new technology to make this training accessible to all public servants for free.
Now New Jersey is partnering with four other states (California, Colorado, Maine and Pennsylvania) in a nonprofit consortium called InnovateUS to expand access to free upskilling for public servants. Governed by leaders from those states, InnovateUS delivers biweekly workshops to public-service professionals from more than two dozen fields and 30 states on such topics as “what the heck is ChatGPT?” with former USAID Chief Innovation Officer Alexis Bonnell to “how to write a generative AI policy” with Boston Chief Technology Officer Santiago Garces.
InnovateUS also offers at-your-own-pace courses in “Innovation Skills for Public Professionals,” which covers how to use technology, data and community wisdom to improve the delivery of services, and “Open Justice,” which addresses how to use these skills to improve legal institutions. InnovateUS has a course planned for this fall on “AI for Public Professionals,” and has already released an explainer video on generative AI for public servants in Spanish and English. The explainer video itself was made using generative AI!
Other states are also taking steps to upskill their workforces on the use of innovative technology. Indiana has trained some 1,800 state employees in data literacy through its Data Proficiency Program. Short lessons explain data concepts with simple, real-life scenarios. Similarly, former California Chief Data Officer (CDO) Joy Bonaguro established a state Data Academy, building on her experience training public servants at the local level when she was San Francisco’s CDO.
The issue programs like those address is that while there is a great deal of content online about technology, it is generally written for technology professionals or Silicon Valley insiders. There is a need for expanded technology education for those who do not “speak tech” but who are interested in uses of technology for public service.
The need for upskilling public professionals in a form and format useful to those who govern is urgent. Of 75 people surveyed who took InnovateUS’ online course on innovation skills, only 13.5 percent said they had prior training in topics such as data analytical thinking or human-centered design. As pioneering American computer scientist Alan Kay famously said, "the best way to predict the future is to invent it." If we want to invent a future where AI improves government and strengthens democracy, then we need to start by learning what it is.
Governing's opinion columns reflect the views of their authors and not necessarily those of Governing's editors or management.
Related Content
