Should priests use A.I. for homily help? A Silicon Valley pastor explains how and how not to

“If you don’t spend time writing your homily today and you just get up and give it, ask ChatGPT or Claude—they’re gonna do a better job than you,” Father Brendan McGuire says. In the third episode of our series on Preaching for Missionary Discipleship, host Ricardo da Silva, S.J., welcomes Father McGuire, pastor of St. Simon Parish in Los Altos, Calif., and a former Silicon Valley technology executive who now advises the Vatican and Anthropic on A.I. ethics. A longtime teacher of homiletics, Father McGuire recently led formation sessions for Jesuit scholastics in Florence on using A.I. in homily prep—and brings that teaching here, making the case for a disciplined “deep prompt,” built from a preacher’s own best homilies, over a “thin prompt” that produces something generic and derivative.
Timecodes:
0:00 AI may prepare a better homily than you
2:29 Meet the Silicon Valley priest advising Anthropic and the Vatican
4:00 Recalling his first homilies — 30 years ago
8:20 Preaching as an ‘art’
11:41 Preaching that leads to “missionary discipleship”
15:33 Not just the words of the homily — but the delivery
18:20 A nun challenges him to change how he preaches
23:00 AI homilies: Yes or No? 25:19
25:58 AI homilies: thin prompts vs. deep prompts
30:10 AI Hallucination and scriptural analysis
37:20 What are the ethical guardrails for an AI homily?
Links from the show:
Father Brendan McGuire’s homilies

Facts Only

* Father Brendan McGuire is the pastor of St. Simon Parish in Los Altos, California.
* Father McGuire is a former Silicon Valley technology executive who advises the Vatican and Anthropic on AI ethics.
* Father McGuire led formation sessions for Jesuit scholastics in Florence regarding using AI in homily preparation.
* The discussion covers the creation of homilies using ChatGPT or Claude.
* Father McGuire advocates for a "deep prompt" built from personal homilies over a "thin prompt."
* Topics discussed include preaching as an art, delivery, and achieving missionary discipleship.
* Concerns raised include AI hallucination and scriptural analysis in homily creation.
* The discussion addresses the need for ethical guardrails for AI homilies.

Executive Summary

A conversation between Father Brendan McGuire, a pastor and former Silicon Valley executive advising on AI ethics, and Ricardo da Silva, a host, explores the use of Artificial Intelligence in homily preparation. Father McGuire advocates for a disciplined approach to creating homilies by using "deep prompts" built from personal best examples, contrasting this with "thin prompts" that yield generic content. The discussion touches upon preaching as an art, the importance of delivery, and applying these concepts to achieving "missionary discipleship." Key topics covered include the potential of AI for homily creation, the risks associated with AI hallucination and scriptural analysis, the need for ethical guardrails, and the difference between superficial prompting methods.

Full Take

The dialogue pivots on balancing technological capability with established artistic and spiritual disciplines. The tension exists between the efficiency offered by AI tools and the qualitative demands of pastoral communication, specifically regarding "missionary discipleship." The framework presented suggests that simply automating content generation ("thin prompts") risks producing derivative material, whereas embedding deep experience into the prompt ("deep prompt") supports a more authentic, artful form of preaching. This mirrors a broader tension in technology adoption: the conflict between scalable, easily accessible output and depth, context, and intentionality. The inclusion of AI ethics advisors suggests a recognition that technological deployment requires not just technical competence but also moral and theological accountability. The implications point toward questioning whether the pursuit of efficiency inherently compromises the reflective depth necessary for meaningful spiritual instruction. What frameworks must be established to ensure that tools designed for efficiency support, rather than undermine, the development of authentic human intellectual and spiritual agency in communication?

Sentinel — Human

Confidence

This text appears to be a transcription or summary of an organized interview, showing the typical structure and flow associated with human journalistic reporting rather than purely synthetic generation.

Signals Detected

Sentence length variance is natural, reflecting interview/discussion structure.

The text flows logically as a summary of an interview, attributing specific points directly to the cited figures and concepts.

No obvious template matching; the structure is driven by the dialogue rather than pre-existing arguments.

The context involves specific religious and academic figures (Jesuits, Vatican, Anthropic) suggesting a real, albeit niche, journalistic source.

Human Indicators

The inclusion of specific names, titles, and the explicit citation of interview structure strongly suggests human sourcing rather than generalized AI summarization.

The nuanced juxtaposition of religious practice (homily) with advanced technology ethics indicates a voice that synthesizes disparate concepts thoughtfully.