I live by Ronald Reagan’s mantra: trust but verify. It’s guided me through decades of challenging powerful interests—from Big Pharma’s push for disease management over true wellness, to flawed nutritional science, and now, even AI tools handling sacred Catholic texts.
My background isn’t just theoretical. In my early days, I cut my teeth on computer programming, where the iron rule was “garbage in, garbage out.” Later, I spent decades building websites, blogs, and video blogs. As a pharmacist still in school, I launched one of the very first wellness sites promoting lifestyle changes, nutrition, and supplementation as alternatives to Big Pharma’s agenda. I’ve spoken out publicly—through websites, vlogs, radio appearances (including guest spots on KSCO Radio in Santa Cruz), speaking engagements, and podcasting—against medical misinformation from the medical-industrial complex.
That activism cost me. I was an early YouTube influencer earning income from my videos until I shared a radio interview titled “What Big Pharma Doesn’t Want You to Know.” In it, I highlighted how Ancel Keys’ seminal research on fats and cardiovascular disease was fatally flawed: Keys selectively published data supporting his hypothesis that saturated fats caused heart problems, while ignoring contradictory evidence showing no link. That exposure led to my channel’s demonetization.
On the faith side, marrying outside the Catholic Church brought sharp criticism from my wife’s family. As a cradle Catholic, I’d never encountered their objections before. This prompted deep study of the early Church Fathers to understand pre-Reformation teachings. When a fundamentalist later challenged me with 18 pointed questions aimed at shaking my faith, it only strengthened my resolve to seek truth rigorously.
A professor in pharmacy school taught me a vital framework for evaluating contradictory studies: When overwhelming evidence points one way but an outlier study disagrees, ask three questions—
- Who sponsored it?
- What do they stand to gain or lose?
- What methods directed the outcome toward the desired result?
I apply this (along with “trust but verify”) to everything: Ancel Keys’ selective data, Covid-era dismissals of natural approaches in favor of vaccination mandates, and now AI responses on religious matters.
The Incident: Testing AI on Well-Known Readings
On January 21, 2026, I queried Microsoft Copilot (Copilot 365) to summarize the Catholic Mass readings for Sunday, January 25, 2026—the Third Sunday in Ordinary Time, Year A. I already knew them from official sources like the USCCB website, but I was using AI (as I do with multiple tools) to help organize thoughts for reflection or writing.
Copilot failed—five times in succession. Each response gave wildly different, incorrect readings, mixing cycles (A, B, C), weekday feasts, or prioritizing the Feast of the Conversion of St. Paul (January 25) over Sunday liturgy, which always takes precedence.
I challenged it: “Explain why 5 times you gave incorrect information. These readings are well publicized by the Catholic Church. A simple Google search gives the correct answer?”
Copilot admitted errors: over-reliance on “internal knowledge” without verification, confusion over cycles and feast precedence, and failure to perform a live check. It finally provided the correct ones:
First Reading: Isaiah 8:23—9:3
Responsorial Psalm: Psalm 27:1, 4, 13–14
Second Reading: 1 Corinthians 1:10–13, 17
Gospel: Matthew 4:12–23
Theme: Christ the Light Who Calls Us to Unity and Mission—Jesus fulfilling Isaiah’s prophecy of light in darkness, Paul urging unity, and the call to discipleship.
Pressing further, I noted that “stored liturgical patterns” and repeated failures sounded like deeper issues. Why didn’t it validate against authoritative sources? Grok and Google got it right immediately. Copilot insisted no agenda—just logic and verification mistakes.
My final reply: “Unfortunately I no longer trust that I am getting truthful answers.”
Broader Implications: Trust, Verification, and Technology
This wasn’t isolated. Like flawed studies on statins or saturated fats—often sponsored by interests favoring pharmaceuticals over nutrition—I approach AI outputs skeptically. Who “programs” or curates the training data? What biases or gaps exist? Why prioritize “internal knowledge” over real-time authoritative checks?
In faith matters, accuracy is non-negotiable. These readings speak of light overcoming darkness—a fitting metaphor for navigating unreliable tech. AI can assist in formulating ideas, but it demands verification against primary sources (USCCB, Vatican). Just as I’ve long urged cross-checking medical claims, I now urge the same for spiritual ones.
Technology serves faith and wellness, but never replaces discernment. For January 25, 2026, verify the readings yourself at the USCCB: https://bible.usccb.org/bible/readings/012526.cfm. Trust—but always verify.
Keith is a Louisville, Kentucky resident, pharmacist, early web pioneer, wellness advocate, and lifelong Catholic who applies rigorous scrutiny to sources in medicine, nutrition, and faith.
Hashtags / Labels for Blogger: Catholic Mass Readings, Third Sunday Ordinary Time, Year A Lectionary, AI and Faith, Microsoft Copilot Errors, Trust But Verify, Ronald Reagan, Big Pharma Critique, Ancel Keys Research, Statin Drugs Truth, Early Church Fathers, Catholic Apologetics, Natural Wellness, Pharmacy Insights, Medical Misinformation, Faith and Technology, Louisville Catholic Blogger
Posts
No comments:
Post a Comment