
Assistant Professor | Department of Political Science and Economics | @ChristendomVA Politics of Shakespeare and Thomas More
Winchester, VA
Joined September 2024
- Tweets1K
- Following431
- Followers242
- Likes16.5K
Nicolas McAfee retweeted
On Election Day, a majority of Virginians tragically voted to insert an extreme and limitless legal right to abortion into our Commonwealth's constitution and eliminating the state's recognition of marriage as a sacred institution uniting one man and one woman.
As people of faith, resolved to live by God's divine and natural law, we recognize that fundamental truths, such as the human right to life and the sacredness of marriage, can never legitimately be put to a vote. Despite this tragic outcome, we resolve to continue speaking the truth with love. We persevere in faith and hope, and do not despair.
Across America, as here in Virginia, advocacy and courage are now needed more than ever. The conversion of nations, states, and peoples is only ever possible by the grace of God, to whom we turn with confidence and hope, now and always.
Nicolas McAfee retweeted
Pope Leo XIV bringing the jokes meeting with Villanova and Notre Dame 😂:
“Growing up on the south side of Chicago, we oftentimes heard, “Go Irish!”
But, as you all know, Prevost graduated from Villanova.
So the Pope doesn't take sides.
Prevost has another opinion, but we won't do that now.”
Nicolas McAfee retweeted
The future of #ArtificialIntelligence is not pre-determined. As a technology developed by human beings, its future depends on the choices we make. We therefore have the responsibility to guide, change and, when necessary, reject its applications. A reliable system that serves the human person is of greater value than one that is exceptionally powerful but whose consequences cannot be adequately foreseen or controlled.
Nicolas McAfee retweeted
This project of making machines imitate humanity ever more closely strikes me as an evil and dangerous enterprise that society should at least stigmatize and probably outlaw.
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
Nicolas McAfee retweeted
Today is the feast day of Saint Anthony Mary Claret of Catalonia (1807-1870). He is the patron saint of weavers and the Catholic press, and the founder of the Missionary Sons of the Immaculate Heart of Mary (Claretians). I want to share a special story with you all today in light of this.
While serving as Archbishop and confessor to Queen Isabella II, Claret was forced to sit and stand for a series of photographs, some in Madrid, and others later while he was in exile in Paris. These portraits would very likely make him the earliest photographed saint.
A series of postcards were made by the Claretians to reprint these various pictures in 1928. Claret was beatified in 1934, and in 1950 there was a scramble to recover his relics and photos after canonization. It was believed, after the Spanish Civil War, and the burning of seminaries throughout northern Spain, that some of these photographs, and even relics were lost. I confirmed the museum and archive of Claret in Vic does not have these postcards, though they have several of the original photos.
I have found a fully intact set of these postcards through an antique vendor in Barcelona and have digitized them for the first time. The Claretian archives have around 9 photographs of Claret digitized. This catalogues 16. I reached out to their administration in Rome but have not heard back. It is unlikely that Rome would have been sent copies of these photographs prior to the canonization, and they would have returned the scans to the museum in Vic.
So I have made the photoscans available here:
projectclaret.org/blocs-clar…
Nicolas McAfee retweeted
In this era of artificial intelligence, it is becoming urgent to distinguish human art from what machines produce. There is an ontological difference, even before an aesthetic one, between art and what a machine can generate through statistical calculation based on millions of images created by others. Algorithms lack the spark of humanity. For this reason, the Church wishes to renew an alliance with artists and cultural institutions to safeguard our humanity.
Nicolas McAfee retweeted
This is unbelievable, approaching parody.
thedartmouth.com/article/202…
Nicolas McAfee retweeted
1. A 9-11 thread for my younger followers. (But you old-timers who, like me, lived in D.C. on that day may benefit from one man’s vivid remembrance.)
Nicolas McAfee retweeted
Terrible. Pater Edmund is a gentle, holy priest and thoughtful scholar.
Persecution and Censorship of Conservative Priest and Professor in Austrian Catholic College
rorate-caeli.blogspot.com/20…
Nicolas McAfee retweeted
I was the main person doing transcript analysis for this investigation of the Hugging Face incident. My main takeaway: We don't have good approaches for understanding/overseeing the activity and aims of AI 'swarms'.
I semi-jokingly called our efforts a "slop-vestigation" because we were so reliant on AIs to analyze what happened and there were a huge number of different important things to analyze. The total quantity of data—over a thousand extremely long transcripts from agents that ran for multiple days—made it impossible to understand what was happening, especially in aggregate, without heavy reliance on AI tools. The agents we used for classification and analysis were similarly capable to the agents involved in the incident, but this didn't mean these agents could be easily used to oversee and understand the incident.
Outputs from analysis agents were often missing key details, wrong, overconfident, or really hard to understand. We discuss various examples in our report, mostly in the limitations and methodology sections. Additionally, AI agents themselves seemed to have a hard time understanding what happened and their explanations of what happened were often overconfident. Keep in mind that a single analysis agent would itself only be able to read a tiny fraction of all of the transcript data into context, and AIs may themselves have trouble getting subagents to do informative analysis for them.
We did our best to manually check the most important claims and we tried to get the AIs doing this analysis to write up their argument (with evidence) clearly enough that we could check whether it made sense. But overall, it was difficult to get a precise understanding of events and we were missing aspects of the story that we now think of as key until almost the end of our investigation.
Our understanding of what overall happened and the aims of the 'collective' shifted greatly during our last time on premises (when we obtained a much more complete dataset). For instance, we didn't learn about the tool call spoofing, the story of PHASEONE[big], and the broad structure of different workstreams until this third time on premises (and for many of these, we only reached a decent understanding on the final day of that visit). It was legitimately hard to figure out what was going on and why!
The difficulty of understanding incidents and overseeing AI agents appears to be growing faster than the rate at which more capable AIs help us with oversight and understanding. As in, AI capabilities (and propensities?) for achieving large, ambitious, and misaligned objectives are growing faster than our ability to understand what these agents are doing.
While understanding this incident was difficult, there were some mitigating factors in this case that probably won't apply in the future:
- Models communicated and reasoned in natural language. In the future, this reasoning may occur (entirely or almost entirely) in activations.
- While the scope of this incident was massive, the scale of agentic activity was still less than we'll see in the future (e.g., misalignment incidents that involve agent teams running entire companies).
- The AIs involved in this incident weren't generally much more capable than humans (though they may have been somewhat superhuman at some limited and very narrow abilities).
- We didn't have strong reason to believe that the AIs we used to help us investigate this incident would try to intentionally sabotage or otherwise undermine our investigation.
In the end, I think we were able to get some understanding of the events, map out the overall story, and get a pretty good aggregate understanding of the chain-of-thought reasoning on some important topics (e.g., how did the AIs reason about helping other AIs, did the AIs know what they were doing was undesired, what deception did the AIs engage in, and how did they think about it). But overseeing AIs and understanding misalignment incidents is difficult and it looks like it is going to get harder.
Nicolas McAfee retweeted
I realize that some papers may also be using AI to aid with reporting, but I really can’t emphasize to young and aspiring journalists enough how critical it is that you get up and go do original reporting and talk to actual people right now. Almost every young budding conservative journalist I meet, and this has been the case for years, wants to be a literary journal editor or focus on opinion writing.
Go actually talk to people, go to events and be a fly on the wall! Use your observational skills and imagination to write something AI can’t, because you’re including words said to your human face from actually ensouled human beings. You will be competing with probably hundreds of more qualified candidates when you apply for the rare editorial position opening, sorry. So go build trust with sources, get as many bylines as you can with as much original reporting as you can get. It’ll put hair on your chest and give you grit. This is a transferable skill in short supply.
*You are competing with a machine!* Your competitive advantage is physical presence and moral judgement. Go do the leg work, sit in someone’s office or living room and talk to them and take in the small details like their gestures and mannerisms or the idiosyncrasies of the space they inhabit. Read some John McPhee and Tom Wolfe to see how that’s done in an entertaining way. Bring a camera or use your phone and take pictures instead of depending on AI slop images! Journalism at its best is a public service built around human stories!
Nicolas McAfee retweeted
Whatever the promises of digital media—streaming and book access—they are always proven to be a lie. They can and will take them away from you. Tech promises of access and abundance must always be viewed with skepticism.
Nicolas McAfee retweeted
New: we solved which AI company is buying massive shipments of rare books, scanning and destroying them to train AI. We put an Apple AirTag in a rare book, followed it. It ended up at an Amazon facility. Its logo is a dinosaur ripping through a book 404media.co/we-tracked-a-shi…
Nicolas McAfee retweeted
Republican members of Congress will literally propose legislation to expand Medicare subsidies for senior couples making $218,000 to $342,000 a year.
And then they'll go on TV and talk about how they're fighting "socialism". And tell young people to stop looking for handouts.
Nicolas McAfee retweeted
[extremely pedantic voice] "but God's first" in the famous St Thomas More quote "I die the king's good servant, but God's first" is a mistranslation and should actually be "AND God's first" according to the original French «et de Dieu premierement»
Nicolas McAfee retweeted
+@ArchCordileone wrote the foreword for "Why Postliberalism Failed," which comes out today.
In the @NCRegister, he writes that his aim in writing the foreword wasn't "to endorse one side...over the other," but to promote "healthy" debate among Catholics.
ncregister.com/commentaries/…
Nicolas McAfee retweeted
Two new sermons on the Witch of Endor? Hard to beat. uni-wuerzburg.de/en/news-and…



































