/Genevan Commons and the Discernment Deepfake

Genevan Commons and the Discernment Deepfake

Steven Wedgeworth

I was recently doxed and deepfaked. If you don’t know what that means, I will explain. For now, let’s just say that it doesn’t make for a good weekend.

A new blog was launched which aimed to expose a secret group of misogynistic pastors and elders. It consists of screenshots, illicitly taken from a private Facebook group called Genevan Commons. The sensational comments nearly all appear within running threads, threads with many participants and many comments. And truly, some of the comments were indeed offensive. Several showed immature jesting, and some went further, involving cruelty and authoritarianism. The comments which fit these categories are absolutely sinful and require repentance.

My initial reaction was grief for the women who had been named and mistreated in the comments, as well as all of the other women who would be hurt by these screenshots. I cannot imagine the depth of pain that this all caused. Then I reflected upon my own inclusion in the group. “Did I really see all that stuff?” I asked myself. “Did I laugh along with all that?” My anxiety spiked. I soon began to receive various upset messages. Most of these, it has to be said, were entirely gracious and sincere. My experience was not like that of some others listed on the dox site, who report receiving abuse and even outright threats. No, my comments were people expressing their disappointment in my behavior and letting me know that I had really let them down. So my desire was to do some soul-searching and then decide what I needed to repent of and how to do so in the appropriate way. I went to the Lord in prayer and sought counsel among trusted pastors for several days. One of these asked me to walk through as many of the details as possible. “You have to be specific for repentance to be real,” he said. After I laid out as much as I could, he gave me this reply:

“Brother, you’ve been deepfaked.”

James Lindsay is a classical liberal writer who studies how certain academic methods in the social sciences are actually used as weapons against truth. In a May 2020 essay, he explains the concept of “deepfake methodology.” The term “deepfake” itself refers to video and audio manipulation, the ability to construct a video of an actual person, with their face and in their voice, saying things that they did not say. This often involves a technology which captures their words and then rearranges them in such a way to make it appear that an actual real event is being replayed rather than a synthetic reconstruction. Lindsay explains it this way:

[A] deepfake is something like an extension of a doctored photograph to include audio and especially video, which is only really possible using advanced machine-learning algorithms and fairly powerful computers. In the audio version, a deepfake will take snippets of a person’s voice and reconstruct how they speak, enabling the production of an audio clip that has them saying anything the creator wants. With video, it’s even more insidious, combining both the audio feature and the ability to replicate one’s face, facial movement, and facial expressions in a video setting.

Some popular deepfakes are simply entertaining, portraying celebrities saying outlandish or absurd things. But other deepfakes are malicious. The internet can now make people say, and do, literally anything. What makes deepfakes so powerful is that the average audience cannot know that what they are seeing isn’t real. They are faked out in the deepest way possible.

But as Lindsay explains, this same methodology can be applied to things other than video. He summarizes a recent controversy within the world of critical theory where two scholars employed this concept of “deepfake methodology” to explain how other scholars had manipulated a literary source. Lindsay quotes from the article making this argument, stating:

[The authors] generally use quotations radically out of context. They never study what is done in the texts they are “reading”. They say nothing about their own methodology or data selection and give no principles for interpretation. They do not define racism… and they don’t discuss at all what it means to read a theory and judge whether it is racist… If there is a methodology at play, it is deepfake in the sense that if you break a corpus of text down into small fragments, you can reassemble it to say anything you want.

Lindsey then summarizes the implications of this sort of thing:

Thus it is that activists, liberated from any obligation to truth and falsity and seeing objectivity as a myth (that’s used to maintain oppression, even), developed and deployed a deepfake methodology: “making somebody ‘speak’ by using splinters from them reassembled to produce meaning disconnected from the original texts.”

This methodology is essential to understand because the recent discernment blog claiming to expose the secrets of the Genevan Commons relies largely on this kind of deepfake. It chops up fragmentary statements from various people and then arranges them in a particular way in order to make them say what the discernment blogger wants them to say. This is a methodology which can very easily manipulate sincere onlookers, naturally appealing to their sense of decency and compassion and inviting them to defend the weak against those on the attack. It invites a reactionary response, and it can be extremely effective. The explanation that follows will not attempt to fully vindicate Genevan Commons. There have been several legitimate expressions of concern in the wake of all of this, and I will try to give my own thoughts about the moral and spiritual aspects below. But before we can do that, we also have to understand how many of us have been manipulated. And we need to see how it happened.

My situation was not a case of altered video and audio. It was not a deepfake in that ordinary sense. But it was nevertheless a deepfake of the kind just described by Lindsay, a case of deepfake methodology. To begin to see this, consider the challenge of adequately explaining what “context” actually means in online discourse.

Genevan Commons was a Facebook group. It was “private,” according to Facebook’s settings, which means that it was not open to the whole world, but it included over a thousand members for most of the time that I participated in it. The membership was originally diverse, and it discussed all sorts of topics. It was centered around Reformed theology, but like so many Facebook groups it could go in any direction at any given time. Over time, it became increasingly clear to me that an unhealthy ethos and tone was dominating the group. This was a gradual process, and some of my friends recognized it before I did. I “unfollowed” the group (which means I no longer saw its content) sometime in June of 2019, and I fully “left” (which means I deleted my membership) in early 2020, as a part of a general Facebook cleanup decision. I had not realized just how many groups I had joined or been added to over the years, and it was clear that most of them were unhelpful.

This explanation is necessary because, for most people, Facebook groups are not secret clubs with membership vows. Many people are even surprised to learn what groups they hold “membership” in. For a season at least, Facebook allowed for unilateral additions to groups. Members could add new members, and the new person did not even have to click a button. It’s also more than possible to participate in a comments thread without reading all of the other comments. Just check any busy page with lots of members. You’ll only be able to see two or perhaps three threads in a single screen view, and when you click on the “comments,” you will likely only see four or five comments, as Facebook collapses the rest. To see every comment, you have to make a point to click all of the “see more” or “view previous replies…” links and then scroll through dense comboxes:

For groups which put out a few thousand comments a day, most comments go unseen. This is simply how ordinary people use Facebook. They skim. They tune out noise. They learn rules like “Don’t feed the trolls.” In a group of twenty people, you can consider everyone a friend. In a group of a thousand, you hardly know most of the participants.

So right away, the framing of a Facebook group as being “complementarians behind closed doors,” as Christianity Today’s Ed Stetzer put it, is already rather fake. Such a description just doesn’t fit how social media actually works. But there are at least five more ways in which the Genevan Commons exposé is a deepfake.

1. The organization of the evidence

The damaging material from Genevan Commons is featured in condensed blog forms and also in a lengthy archive. But this evidence is not organized in the way that it would have been in the actual Facebook group. It does not state the original post (the “OP”). Instead, it is organized according to shocking catch phrases. This gives the impression that the catch phrase represents the topic or thread on GC, the thing in which all of the commentators were participating. But in nearly every case, the problematic comment appears in the middle of a lengthy thread. Some people may join in that side trail, but most do not. One could question whether certain less offensive comments smoothed the way for greater offenses to be tolerated, and I am certainly asking myself those questions today, but this is a truth more easily known after the fact. Earlier commentators cannot be judged simply on the basis of later comments left by other people. Still, for the reader of the exposé to even see any of this relevant information, the information which would explain who actually said what and how, they must first pass through the “problematic” gate set up by the framing of the website. They have been conditioned to see a certain thing and to interpret all of what they see according to that interpretative lens.

2. The “data dump” model

In addition to the misleading framing of the evidence, there is no sincere attempt at interpretation. Indeed, the need for interpretation is assumed to be unnecessary, since the screenshots are there for all to see. But the screenshots are provided in bulk, in a sort of data dump. The quantity of comments is meant to shock, and the header provided by the site’s editor, or the summary explanation provided by various online promoters, is meant to suffice. This is similar to a headline which does not accurately reflect the body of an article. Ordinary readers will accept this framing and assume that most of the many comments are more or less alike.

3. What’s left out

One of the most deceptive aspects of this project is what is not present. As explained above, it foregrounds selective content (some of which is truly objectionable) and downplays or omits entirely everything else that was discussed, both in topic and tenor. It makes the Facebook group appear to always have been devoted to one or two topics, and it almost never includes those occasions when participants did indeed call out abuses or bad behavior. Of course, by its very nature, guerrilla journalism cannot know what was discussed in more private ways. It cannot know if one member of the group contacted another member of the group directly, nor if they wrote to the moderator. Further, it cannot capture what was deleted from the group, perhaps at the request of some of the group members. The combination of sensationalizing some content and omitting other content gives an entirely dishonest picture.

4. Many of the screenshots are rearranged

As I was checking my own comments in the various screenshots, I noticed that some of them had been manipulated. A few of my comments were presented out of chronological context. In one case, they appear after someone else makes a more unsavory or unkind comment, thus implying that I am both excusing that comment and joining in a sort of dog pile. But if you examine the timestamp, you can see that my comment was made prior to the other, even in a different thread entirely. The deception is even worse when one looks closely at the “teaser” or advertisement pictures used on Twitter. Many of the comments there, most of which do not involve me, are actually floating around in arranged shapes.

This means that those comments were cut out of their original space and then pasted on to a new graphic. They may or may not actually have occurred together. And some screenshots have been cropped. This technique was used to leverage some men’s names against others. In the above picture, Dr. Mark Jones is simply asking a question about the original post, yet he is positioned directly across from someone else who is making a demeaning joke. This gives the impression that Jones was also participating in the joke, which is not the case.

One example of cropping concerns a ministerial student named Zack Groff. He is shown to be saying “What happened to her face?” Any ordinary reader would assume that this is a comment where Groff is disparaging a woman’s physical appearance, a truly despicable thing. But in the original screenshot, it can be seen that he is talking about a cartoon illustration, a picture of a toy fairy, and a toy which does indeed have some sort of damage on its face. The original meaning is changed entirely by the cropping, and it does so in a way that certainly seems malicious. In addition to this, at least one set of screenshots was taken from a different Facebook location. The comments were not even made in the Genevan Commons, even though the discernment blog gives every indication that they were.

5. Some screenshots are simply airbrushed

I cannot say how many of the pictures have been blatantly airbrushed, but one such case involves me. This was my first clue that something truly disingenuous was going on. In one screenshot, my face and comment are hanging in the midst of an odd amount of white space. I am pictured saying “literally no expectations.” There’s a blank above me, and there’s a blank below me, and my comment is off center. This is not how Facebook threads normally look.

In the edited form, I am made out to be making a comment directly about Aimee Byrd, that I have no expectations of her work (perhaps because I am a chauvinist?). To make matters worse, I am paired up with other comments which look to be making fun of her. So it appears as if I am demeaning Byrd.

But if the reader does the more demanding work of looking up the unredacted original, buried under a few layers of links, things become substantially different. The blank space was actually once filled in, and by quite a bit. Another commentator appeared above me, and he had quoted a line from a book endorsement which said “unfettered by social and historical expectations.”

With the full image, it is plain that I was not talking about Byrd. I was directly replying to this comment about expectations. My use of “literally no expectations” was meant to indicate that the statement sounds hopelessly naive. Everyone has social and historical expectations.

This was so obvious that the makers of the discernment blog had to graphically alter the picture. They actually deleted necessary content in order to deceive their audience. And the fact that certain commentators were airbrushed out in order to emphasize my comment leads me to believe I was targeted. Whoever made this site has an interest in causing me harm.

It would not surprise me to discover that there are more such cases of manipulation and deepfaking in the various screenshots. What I have seen is enough to convince me that readers cannot form an honest and just conclusion about the various commentators, nor the group as a whole, from the material that is provided. The evidence is significantly skewed. Some of this may have been due to the discernment blogger’s own misunderstanding, but in some cases, it is difficult to avoid the conclusion that the modification was calculated and intended to do maximum damage.

On this point, it should be noted that two tortious acts have likely been committed in this project, namely intrusion of seclusion and false light. My own principles instruct me to avoid litigation in matters having to do with the faith, but it should be noted that the underlying principles of 1 Cor. 6:1–11 have already been violated by the leaker. Anyone currently publishing the Genevan Commons screenshots, and those with any intent to release similar sorts of leaks on others, should consider their own liability here. I highly doubt that everyone shares my position on lawsuits.

None of what I have written above is meant to give Genevan Commons a mere clean bill of health. Once the distorting effects of the deepfake are removed, as best as they can be, there still appear to be a number of commentators who engaged in unbecoming and sinful rhetorical behavior. All who did so should repent. But the number of explicit sins of this kind are in fact a comparatively small subset of the larger number — a thousand members who were participating in a group always in motion, and a group which changed over time. It is impossible to know who saw which comments and who did or did not speak out against the abusive ones. The fact that deepfake methodology was employed suggests that this discernment project was not a simple quest for truth and justice, much less a preliminary action in a process seeking repentance, forgiveness, and reconciliation. It was itself a malicious activity. When you combine it with the element of doxing that was also used (see here and here), the total product becomes a case of cyberbullying in its own right. Doctored images were presented in a way that intimidated and harmed many innocent people. The doxing of every member’s name and place of employment threatened their livelihoods and their families’ wellbeing.

For my own part, I do indeed regret the levity with which I participated in discussions which should have been much more serious. In ordinary situations, a person would simply repent and not try to explain away what happened. All that I have written above would be inappropriate and unnecessary. But in this situation, an ordinary response would actually be quite irresponsible. It would be akin to a forced confession within a larger corrupt process. It would give credibility to accusations which are meant to harm others, including entirely innocent people, as well as the various institutions which have associational connections to those people.

There are more than a few lessons to be learned here. The first is to learn the perils of online discernment, how people use and misuse the internet and how the internet manipulates us. Christians certainly do need much greater discernment here. We need to be better writers and better readers of what we encounter, and we need to be much slower to speak. I have resolved to quit all Facebook groups and to try my best to cease from all “recreational” apologetics; these things have the appearance of righteousness but are too easily a cover for the sins of the flesh. Many pastors should probably hide or delete their social media accounts entirely. But, importantly, not all should. We do need good men out there, writing good content and rightly standing against the forces which seek to redefine our faith and practice, especially those who do so in very subtle ways. In all of this, we should absolutely resist the temptation to indulge online toxicity, especially those more trendy and enticing varieties which make it appear attractive.

At the same time, we need far fewer discernment blogs. As readers, we must train ourselves to be much slower to react to sensational stories and learn to see how they too play to our baser impulses and desires. The first breaking headline is frequently incorrect, if not fake entirely. Indeed, there’s something satanic about the call-out culture of the discernment blogger. While taking up the mantle of justice, they are actually playing the part of the accuser only, taking their case to the court of public opinion rather than any orderly courts of adjudication. They trample over all standards of evidence and rules of procedure. Further, they invite others to take vengeance into their own hands. This makes true repentance more difficult, and it leaves behind a train of bodies and rubble — new collateral damage for a new cause. Those who feel themselves to have been bullied can quickly become the new bullies, and both teams end up looking more alike than different.

Most of all, moments like these remind us how we all stand condemned before the judgment seat of God. If He should count our iniquities, online or off, none could stand before Him. We are all in need of the grace of Christ for any hope of salvation, and so we had better fall to our knees.

Original Source