For years, we’ve been writing about the importance of good data, based upon which states, counties and cities can make good decisions.
But as a growing number of practitioners and academics are relying on artificial intelligence as a source for all the information in the world (when we were children, the equivalent was the Encyclopedia Brittanica), the necessity of care for the accuracy and completeness of data rises even more in importance
The well-written, if sometimes pedantic prose that AI can produce can create the impression that the information it spits out will be accurate. That’s not the case. Consider, for example, a BBC February report that examined news summaries generated from artificial intelligence (AI) engines including OpenAI’s
ChatGPT, Microsoft’s Copilot, Google’s Gemini, and Perplexity AI.
Here are a few of the report’s alarming findings:
“51% of all AI answers to questions about the news were judged to have significant issues of some form.
19% of AI answers which cited BBC content introduced factual errors – incorrect factual statements, numbers and dates.”
“13% of the quotes sourced from BBC articles were either altered from the original source or not present in the article cited.”
The basic problem is this: as a growing number of people rely on AI to get the information they need, their attention is being diverted from the basics of data management. With so many government officials focused on artificial intelligence, the critical issue of data quality is getting less attention.
A conversation with Rudy de Leon Dinglas, Chief of Staff of the Bloomberg Center for Government Excellence (GovEx) at Johns Hopkins University, revealed his concerns about this issue: “I don’t think we should turn our back on basic data management strategies, because they’re very foundational. AI is here, but we cannot turn our backs on making sure that we’re shoring up the infrastructure about data strategy and data policy.”
Added Kel Wang, manager of applied data practices at GovEx, “All around, everybody talks about AI, but it’s honestly hard to come across articles around the lines of data quality or data inventories or data governance. Data quality is not the shiny icing on the cake.”
Ultimately this goes back to the old adage: garbage in garbage out, a phrase that’s been around since the early days of wide-spread computer use.
We brought up one significant issue with the accuracy of information drawn from AI, in a piece we wrote based on a roundtable discussion about the topic hosted by the IBM Center for the Business of Government. As we wrote, “AI can pick up opinions and understand them as fact, which are then used to make decisions. But the confusion between opinions and fact is a significant one, and when they are conflated, public sector leaders can be misled.”
We’re hardly alone in this concern. The MIT Sloan School of Management had this to say, “AI tools like ChatGPT, Copilot, and Gemini have been found to provide users with fabricated data that appears authentic. These inaccuracies are so common that they’ve earned their own moniker; we refer to them as ‘hallucinations’.
“For an example of how AI hallucinations can play out in the real world, consider the legal case of Mata v. Avianca. In this case, a New York attorney representing a client’s injury claim relied on ChatGPT to conduct his legal research. The federal judge overseeing the suit noted that the opinion contained internal citations and quotes that were nonexistent. Not only did the chatbot make them up, it even stipulated they were available in major legal databases.”
In a recent e-mail conversation with Doug Robinson, executive director of the National Association of State CIOs, he underlined this point, writing that “GenAI models are trained on and use massive amounts of data to be successful. You need to feed the beast! We've identified data quality as one of the most essential elements of GenAI adoption in the states in our blueprint. If you can't trust the quality, integrity and reliability of your data, you can't trust the results of the analysis.”
Last year , NASCIO collaborated with EY on a national survey of the states about this concern, and discovered that “very few states comprehensively address data quality and integrity,” wrote Robinson.
Despite that unfortunate bit of news, the same report indicated that “Ninety-five percent of the respondents believe that increased adoption of AI and generative AI (GenAI) will impact the importance of data management.”
Robert Osmond, Virginia’s Chief Information Officer was referenced in the report as reinforcing the importance of keeping close watch on data quality before allowing AI algorithms to thrust the information into the world, “emphasizing the state’s commitment to ensuring that the foundational data for AI, particularly large language models (LLMs), must be reliable and of high quality. So that AI is ethical, responsible, and transparent, the foundational data used for AI must be accurate, and the results of the AI must be thoroughly tested for accuracy.”
You’d think that the advent of AI, which makes data vastly more accessible would have encouraged governments to focus more closely on cleaning up their data than ever. Sadly, that doesn’t appear to be the case in many places. The glittery prospects of AI are taking up a huge amount of government capacity, and it’s easy to get so lost in the prospects for the future, that the basics of data management that are essential for AI to be useful can lose traction.
#StateandLocalGovernmentData #StateandLocalDataAccuracy #StateandLocalGovernmentGenerativeAIPolicyandManagement #DataAccuracyAndArtificialIntelligence #ArtificialIntelligenceDataAccuracy #DataQuality #StateLocalDataGovernance #CityData #CityDataGovernance #StateLocalGovernmentAIPolicyAndManagement #CityDataQuality #StateandLocalDataQuality #StaeandLocalGovernmentManagement #GovEx #BadData #StateLocalArtificialIntelligenceManagement #CityTechnologyManagement #StateandLocalTechnologyManagement #StateandLocalPublicAdministration #CityPublicAdministration #BloombergCenterForGovernmentExcellence #RudydeLeonDinglas #KelWang #DougRobinson #NASCIO #IBMCenterBusinessOfGovernment #RobertOsmondVirginiaCIO #BandGReport #BarrettandGreeneInc