Buying into AI Mysteries: Worth Every Penny?
This is a revised, updated version of the commentary posted on June 2 at OnWord.net. That blog is now marking its tenth anniversary.
Let’s call this the “Great Penny Mystery”: Has the U.S. Mint ceased producing the beloved, but largely moribund, Lincoln one-cent coin?
I ask this question not because I want to generate breaking news, but because it will help us evaluate human intelligence vs. artificial intelligence. Spoiler alert: I am virtually certain the answer to my query is “no.” But we still can learn something by giving AI a penny for its thoughts.
This puzzle is not inherently difficult. Any Washington journalist whose “beat” includes the Mint or the Treasury Department could contact a public-affairs staff member, receive an answer, and report it.
I am probing a different aspect. I fear that we face a future where our learning could be based less on traditional reporting and more on a credulous mind frame of half-hearted, half-human browsing.
Pope Francis, one of my favorite critics of current tendencies in media technology, scolded journalists in a 2021 message for repeating talking points about people and topics “without ever hitting the streets.” I accept the criticism.
During the past two weeks, without calling any PR person, I used a combination of proactive online inquiries and an AI-assisted search engine to resolve, or perhaps to aggravate, the one-cent conundrum.
I started this with a conventional Google search about the federal government’s use of metals, a topic I had covered diligently in a previous newspaper job. The January 2022 piece at the Market Realist website caught my eye when it reported, “The U.S. Mint announced that it will start to phase out the production of the penny by the end of 2022. The last batch of pennies will be minted on April 1, 2023.”
This supposed development would be big news for several reasons, with implications for everyday business transactions and for the public’s assessment of various economic policies.
On June 3, I went to my trusty AI-equipped search engine and asked the question: Has the U.S. Mint stopped producing pennies?
My aide replied, “Yes, the U.S. Mint announced that it will stop producing pennies beginning in 2022 and will make its last batch on April 1, 2023. The last batch will be sent out in proof sets to be auctioned to coin collectors. Is there anything else you would like to know?”
The primary source cited for this knowledge was the Market Realist piece, along with two other texts, also from 2022.
Ever the skeptical reporter, I chatted further: “Is there proof that the Mint has stopped making pennies?” My aide repeated itself with a touch of impatience. “Yes, there is proof that the U.S. Mint has stopped making pennies. The U.S. Mint announced in early April that it will stop producing pennies beginning in 2022 and will make its last batch on April 1, 2023.”
Sorry, friend, but that’s not proof! I went looking for more recent stories in the mainstream media, for the latest news releases from the Mint, and finally for discussions within one group that I knew would care: coin collectors themselves.
Only then did I hit paydirt. A site called “Coin Update: Daily Coin-Collecting News” published, on May 23, 2023, a lengthy interview with the director of the Mint.
Far down in the story, the interviewer asked about any pending discontinuation of the penny and nickel, noting that the beloved Lincoln cent costs more than twice its face value to produce. Mint Director Ventris Gibson responded that, despite the waste, “there are no plans to discontinue either coin.”
On June 4, I made another foray into “Google News” and found a Washington Times story from this April that focused on the still-valid question of changing the metal content of the penny and other coins in order to save money.
I learned the name of new legislation that would bear watching among those interested in this subject. But the reporter said approval of the bill is iffy.
(I agree: The penny debate is likely to go on, unless or until the government enacts sweeping legislation to transform our currency—which is obviously a scenario worth monitoring.)
On June 11, I returned to my AI chatbot, wondering if it might have digested the updated information I had found: “Has the US Mint stopped producing pennies?”
Sure enough, the browser responded, “According to the U.S. Mint, the agency that makes the nation’s circulating coins, the penny is still in production as of 2023.” That was it, although it added a couple of sentences about the penny’s history and design.
This progress made me even more curious: Had the chatbot enjoyed a thoroughgoing re-education? I asked a slightly altered question: “Has the US Mint stopped producing the penny coin?”
That triggered the stubborn, old answer: “Yes, the U.S. Mint announced that it will stop producing pennies beginning in 2022 and will make its last batch on April 1, 2023.”
This is what prompted me to write with a smile, above in this commentary, that I am virtually certain the answer to my query is “no.” My virtual research aide confounded me, proffering both options as of June 11.
One clue I found regarding this violation of the principle of non-contradiction: It had answered my two questions differently, in one case searching for “US Mint penny production,” while in the other, translating the search into “US Mint stopped producing penny coin.”
We often see this in analog life—the answer we get depends on the question we ask.
All of this is to say that, in my mind, a mystery still exists—but it is mostly about my chatbot, not about the loose change lost under my couch cushions.
We can say these “large-language” AI models—including our present-day systems still in development—are amazing “learners” that can advance society’s knowledge immensely. But, at least for these purposes, they strike me largely as “middle-men” who redistribute words, phraseology, and insights. Their algorithms merely observe materials, in or out of context, and mint shiny new versions.
The systems accomplish their prime directive—to give us answers that sound definitive and persuasive. But this can lull us into a false confidence about their reliability; we assume they share our human kind of curiosity and logic, our desire for orderly inquiry and full truth.
Instead, it seems mega-bots are denizens of a chaotic internet filled with time warps and factual vacuums, eerily adaptive to a post-truth society, better suited to hunters than to gatherers. My friend Steve Dunlop, a former journalist who now consults at the helm of Dunlop Media, recently posted a brief audio report that reveals how he found the ChatGPT bot guilty of errors and “hallucinations” when it wrote an article about him. Listen to his “Does ChatGPT Think You’re Dead?” here.
Truly liberated minds and hearts, as Pope Francis pointed out, should lead journalists to “hit the streets” to find fuller truth. Since most people today play “reportorial” roles by consuming, transmitting, and prioritizing information through social media and podcasts, we all should recognize how our collective pursuit of practical wisdom is different from artificial intelligence.
We are built to ask good questions, to explore and encounter reality as the spirit moves us, to pay heed to the voices and values of people we have not yet heard from. Our wisdom can interact and bond, interpret time and space, unite the natural and the supernatural, and connect data in the “cloud” to unique, healthy tidbits of grass-roots nosiness. Remember, in lieu of further announcements or actions by the Mint, I found coin collectors who handed me a “news update” through a digital back door.
The attributes of virtuous wisdom deserve to be welcomed as part of an unfolding mystery. They leave us extra room to apply creative questions in alignment with time-tested, values-infused answers, all for the sake of a truth which brings meaning, even excitement.
We need to counter and complement AI with a “Star Trek” algorithm. Advanced circuitry may excel at building starships, but only human intelligence will inspire us to explore strange new worlds, to seek out new life and new civilizations. Or at least to look up at the skies with imagination and hope, wondering what we shall do next.
Tomorrow, I will boldly ask my chatbot if it can discern any connection between Captain Kirk and the Great Penny Mystery. This time, I think I’ll have an edge.
Image from ClipSafari.com, a collection of Creative Commons designs.
Keep reading with a 7-day free trial
Subscribe to Phronesis in Pieces to keep reading this post and get 7 days of free access to the full post archives.