No, government use of artificial intelligence won’t save taxpayers money
If history is a gauge, government AI poses a significant threat to liberty
I have no particular desire to throw shade at the Mountain States Policy Center (MSPC), but since I’ve made a point about the free market think tank’s disappointing anti-free market stance regarding milk, here’s another bothersome policy recommendation the think tank made worthy of a critical look:
I only mention it here because it has to do with artificial intelligence and the idea that AI will improve government services and reduce costs. This is highly unlikely to be AI’s function in a government setting. Let’s explore further:
In an April 6 blog post (and repeated as part of the organization’s policy manual), Sebastian Griffin (who also happens to be a Nampa, Idaho city councilman), claims that “AI can improve government efficiency and save taxpayer money.”
No doubt policymakers will try to use “efficiency” and the potential to stretch public money to argue in favor of AI in government. But it won’t work the way Griffin imagines. Technology in government is always deployed under the guise of saving money.
I don’t like to use the word “never” because one generally will find an exception to the rule, but in government, it can safely be said that technology never saves money.
Technology might free up money for other uses, but you can be assured: there will not be fewer employees nor money returned to taxpayers as a result of technology usage, including AI. Will state government agencies fire their communications staff because AI can write press releases? No. Communications staff will just have more time to produce more propaganda. Any “efficiency” AI brings will just be redeployed, freeing government to do the same invasive things as it always has, and perhaps more.
I remember some two decades ago when governments across the United States began deploying “e-tickets” as a means of making traffic enforcement more “efficient” and less costly. But e-tickets didn’t reduce costs; they merely freed time for officers to write more tickets.
In the linked example, a Washington state county went from each of its deputies writing 12-15 tickets a day to writing 18-25 citations a day. Says a report on the matter, “Police forces large and small have reported their e-Citation investments paid for themselves within a year because of the efficiency gains and revenue increases they produced.”
But, says MSPC’s policy manual, AI in government is not all about cost savings. “It 's about delivering better public services and improving the lives of citizens,” the policy manual declares. “By using a good process, public officials can effectively put AI to work for their state.”
Again I remind the reader, government doesn’t exist to improve the lives of citizens. Government exists to protect rights. But alas, government has gotten away from that modest proposition, and therefore, is a fearsome beast.
Now imagine the power of AI being harnessed at, say, state and local tax collection agencies. Government efficiency, powered by AI, being used to subject more taxpayers to more audits. Or AI being used to more readily identify residents who could make use of a government program. Or AI being used to more efficiently argue for more, larger programs because of the efficiency of the analysis of large batches of data about homelessness or hunger or medical status. Or, frighteningly, to augment sophisticated surveillance systems already deployed on our homes, businesses, and cars.
No, the mixture of AI and government will not lead to any real benefit for people. It will make government much more intrusive, and damned more efficient in its intrusiveness.
There’s a Pollyanna-ishness to the idea that AI will have the results MSPC dreams it will. There is significantly more downside than upside for those of us who worry about our freedoms being jeopardized by AI.
Rather than concocting ways AI could be used in government, policymakers should consider AI as a potential threat to the notion of limited government and individual freedom, and they should write policies aimed at mitigating those threats.