Jim Davis and I are giving away something we are proud of - a toolkit to help local church leaders.
This data-driven resource will help your leadership team close the back door, open the front door, & send your members out better equipped.
Brief thread:
https://t.co/pG8ILzCRZO
From my experience the five most powerful words a man can say to another man are:
“I’m proud of you.”
This has proven true in discipleship, parenting, and pastoring.
It’s hard to quantify why, but I think it’s something we need to be telling each other more frequently.
I make a case for why the principled pluralist framework for handling religious questions by LLMs is a significant upgrade to the true/false and pluralist paradigms here in this interview.
We can immediately eliminate religious bias and have religious freedom with this tweak:
The Pope's encyclical on humanity in the age of AI has brought new attention to the relationship between frontier technology and religious belief.
@msgwrites has thought deeply about what this means for people of all faiths and for the leading AI labs themselves. His suggestion is that the default approach for AI models on questions of religion should be one of principled pluralism.
“You’ve asked a question about the Christian tradition. We’re going to answer this question from the perspective of adherents of the historic Christian tradition.”
Then the model gives a full answer from that tradition.
“If you were looking for a skeptical approach, a different religious tradition’s approach, or a sub-denomination’s approach, we can take the conversation in that direction.”
@giskyexplorer@MTSlive In this approach whatever religious tradition is being asked a high altitude question gets to thoroughly share their view & it’s clearly framed as the viewpoint of that tradition and not fact.
The user is always given opportunity to take the conversation in new direction…
I created DO NOT HALLUCINATE today...
One more Substack the world probably doesn't need...
It's free and designed to rapidly bring ministry leaders up to speed on that weeks' biggest developments at the intersection of AI and religion:
@AlmostMedia@pmarca@ryanburge Not a problem at all.
It’s difficult to wait to talk about everything we’ve learned.
This is only data point we’ve shared yet and only because I couldn’t justify keeping it back given the need for more robust conversations on religion and AI.
@AlmostMedia I conducted this study with @ryanburge - the sample size is over N=100,000 and we publish on this in early 2027.
There are complex dynamics of covariate institutional trust and religious attendance.
I elaborate more here in my short manifesto:
https://t.co/yEApToyBnx
AI models are quickly becoming humanity’s "high priests."
We are shifting from a species that acquires knowledge through primary sources to one that relies on secondary AI sources.
This leaves a significant problem.
Here is my manifesto on how to fix it. 🧵
This approach protects models from bias claims, satisfies religious freedom advocates on the right, and pleases pluralism advocates on the left.
Every religion gets to compete fairly in the marketplace of ideas.
These issues are fixable, fast. Let's talk.
My DMs are open.