Liz Fong-Jones / Elliot William Fong / @lizthegrey - 'Consent accident' enjoyer, ex-Google employee, nepotistic sex pest, Robert Z'Dar look-alike who wants authority over the Internet

Strangely, these old tweets came up in a recent Google search I did. Not sure if they've been posted here before, but just in case:
https://web.archive.org/web/20220907041511/http://twitter.com/lizthegrey/status/1567003479285301248 [Archive]

First tweet from "Debi Jackson", who has her own KF thread here.
0.png
Debi Jackson said:
So I turned to Google. I was connected to folks on their LGBTQ ERG and they helped. After a couple of months, they realized how bad KF was and decided to stop indexing their pages. That meant no more KF pages would turn in up search results for any person they targeted!
Then Liz Fong-Jones responded:
1a.png
1b.png
1c.png
1d.png
Liz Fong-Jones said:
Hi! I was one of the people who worked on this.
KiwiFarms stopped turning up in autocomplete for search terms but was still eligible to appear in search results as of 2018.

the rationale is that the google web search team views search autocomplete as more of an editorial decision they can change, but that they're very hesitant to reach in and make tweaks with regard to actual SERPs.

still, it definitely helped a LOT when "liz fong-jones" stopped suggesting "liz fong-jones kiwifarms'

and to be clear, while I was acting on behalf of Trans@Google, the ultimate decision was made by the Google Web Search team and them alone; I exercised no undue influence aside from bringing the issue to their attention through the routine process for unhelpful search behaviour.

the team was very clear with me that if I wanted to pursue de-listing of Kiwifarms for any individual, that a court order eg for defamation would be required, and that the court order would then be posted to http://chillingeffects.org / Lumen, in a sign of hilarious irony.

because nothing says "chilling effect" like having doxxing and stalking lol

There's been progress since then, for instance now individual doxxing pages can be denylisted for individual people's names as web search results, but it's page by page and search term by search term; there is no blanket process to denylist by domain or by search term entirely.

honestly, I'm relieved to have dealt with this problem at the source rather than needing to policy litigate this through the Google Web Search team, who are almost as resistant as Cloudflare when it comes to marketplace of ideas thinking vs harm reduction tradeoffs in SERPs.
tl;dr
Liz Fong-Jones admits that he was the one who got the ball rolling on de-listing Kiwifarms from Google search results.
 
This fucking censorous shitstain. I wonder when timelines mesh for his business of tall poppy and trying to sell reputation cleansing to the wealthy (per MATI and earlier posts) and his attempts to get a stubborn little corner of the internet taken down.

Also lols. Google by 22 was massively manipulating search results and was dead to anyone who gave a shit. That they were once hesitant just shows how far Jewgle and the Pajeets has fallen from any early ideas. (Also thats definitely my new band name)
 
First tweet from "Debi Jackson", who has her own KF thread here.
This got me to check her thread for the first time in forever. It looks like comments ITT about some neighbors mentioning her made Debi join the forum and post some random Spanish woman’s information. (Potentially, of course. Obviously I cannot be certain this is Debi.)

In the Twitter thread posted above she explicitly states that she checks KF.
Now that we've both moved, I guess I don't feel PLish saying I used to walk my dogs past their house most days. I learned about them from other dog walkers before I knew about this thread, as an aside since I know she was afraid of stalkers.
I thought it was weird she always renewed her vanity license plates if she was worried about privacy.
What did the other dog walkers tell you about them?
I think I accidentally got her to join and comment though - solidous10 made an account just to reply to me here and on my profile.
I can assure you I am not Debi Jackson.
Nobody responded to this, the thread is super dead. 9 days later the account came back to post this:
I did some research and found out that Debi and her family have relocated to Torrent, Spain.

Debra Teter nee Rodvelt​

Calle Calderón de la Barca 0
46901 Torrent
Valencia - Comunidad Valenciana - Spain
IMG_8485.jpeg
 
Last edited:
I went looking for a possible recording of Liz Fong-Jones' presentation at the University of Cambridge in November 2023.
Unfortunately, I think Liz probably requested that they not record it. If a recording existed, it should have / would have been archived here: https://www.cl.cam.ac.uk/research/security/seminars/archive/video/
I also found this: https://talks.cam.ac.uk/talk/index/207463
canceled-deleted.png
Liz definitely did the presentation (title: "Reverse engineering hate" - about his feud with Kiwifarms). Sounds like he was worried about potential legal issues at the time, although he later had his talk vetted by a legal team.
vetted.png


Next, I went looking for a recording of Anh V. Vu's presentation of his academic paper (title: "No Easy Way Out: the Effectiveness of Deplatforming an Extremist Forum to Suppress Hate and Harassment" - about the impact of the #DropKiwifarms campaign).
This was presented last month at the IEEE Symposium on Security and Privacy (IEEE S&P) 2024, which is notable since the IEEE is very important/influential. Null talked about it in his Online Censorship article.

Schedule: https://sp2024.ieee-security.org/program.html (Ctrl+F "no easy way out")
ieee-paper.png vu-and-lfj.png
Unfortunately, I hit another dead-end. The youtube channel for IEEE S&P uploaded their 2024 conference presentations just a few days ago: https://www.youtube.com/playlist?list=PL0pRF4xvoD0kKDUYKpKpbOiVipYN2gmS0
...but I don't see Anh V. Vu's presentation? 🤔 I wonder if he requested that it not be posted?



I do have some good news, though. Consolation prizes!

Here are Anh V. Vu's slides for the IEEE presentation:
https://anhvvcs.github.io/static/media/vu2024no-slides.pdf (I'll also attach the pdf to this post)
1.png 2.png 3.png 4.png 5.png 6.png 7.png 8.png 9.png 10.png 11.png 12.png 13.png 14.png 15.png 16.png 17.png 18.png 19.png 20.png 21.png 22.png 23.png 24.png

And here is an audio recording of Anh V. Vu talking about the paper at Harvard University earlier this month:
https://youtu.be/lVKoeuka1h0?list=PLbyW0t9gkXg1cETjigSLl9W3fPKS7Fm8L&t=2164 (Timestamp: 36:04)
This was part of the Workshop on Security and Human Behaviour (SHB'24), held on June 5th.
https://vaniea.com/post/shb-2024-day2/#session7 🙄🙄
wickedness.png
Clip of Anh V. Vu's portion:

TRANSCRIPT:
Anh V. Vu said:
Hi, my name is Anh. I'm a PhD student at Cambridge. I'm going to be talking about a concerted effort by industry to take down hate & harassment forums. This is the work that Jin (?) mentioned yesterday, and this is a joint work with [Alice Hutchings] and [Ross Anderson].

So we published this paper a couple weeks ago at Auckland (??), and Ross* was the person who wanted to talk about this today. I do have a plan to talk about another paper, but I will try my best to do what Ross wanted to do.
So the forum we looked at is Kiwifarms. How many of you have heard about Kiwifarms before? (pause)

Well, half of you? For the rest of you, Kiwifarms is probably the largest forum known for online harassment, where you can also find unpleasant content about cyber bullying, doxings, and a lot of information of users reviewed on that forum.

So this forum has over one decade of lifetime with over 15 million posts in total, and it had been been using Cloudflare protection services for years until September 2022. There was an incident which happened with a Canadian activist in London, Ontario, where it is believed that she was harassed by forum members so seriously by a swatting incident. And then she started a campaign on Twitter in order to pressure Cloudflare and the other tech firms to drop the forum.

And Cloudflare took actions very shortly. So it blocked Kiwifarms, and in their public statement, it is said that they blocked the content because of immediate threat to human life. Then Kiwifarms quickly moved to another service called DDoS-Guard, which is a Russian-based one, but just one day after, DDoS-Guard also dropped the forum. And the forum then moved to another platform, but it got blocked very shortly afterward.

It was not just about three of these tech firms, but many others also tried to block the forum, including some T1 networks. This is very extraordinary, because we have just about fifteen T1 networks in the world, and three or four, four or five of them blocked that forum.

And because this case is very special, we wanted to evaluate how effective this take-down campaign is. Our method is that we collected the entire forum data spanning over a decade and the Telegram channel used by the forum users, because they tend to move to other platforms when the forum was not accessible. And we also collect web traffic to the forums and also the showed interest of Internet users about Kiwifarms.

So here is our first impressions. So, there was an unintended consequence. As you can see here, the showed interest of Internet users and web traffic to the six major domains of Kiwifarms increased dramatically - like 7 times and twice higher than before, just right after Cloudflare's actions. That means more people are more interested in searching Kiwifarms and also to visit the forum.

Here is the forum activity over time, and you can see a very big drop here just right after Cloudflare's actions. And the second big drop was after a T1 network blocked the forum. But eventually, after all, after a couple of months the forum recovered. So this figure is until December 2022 when we submitted the paper. And at this moment, the forum activities fully recovered, and it is even higher than before.

Then we ask the question: Where did the users move to? Well, they moved to Telegram. And here you can see a very big spike of activities on the channels, and you can see the scale here - it was even higher than the forum itself.

Here is the number of postings on Lolcow.farm, which is the primary competitor of Kiwifarms. So we do see some drop here in sync with Kiwifarms, but what interested me - users did not move to that forum. So they tend to move within their ecosystem.

And here is the traffic to six major domains of Kiwifarms. You can see here - the traffic was to the main domain of Kiwifarms, dropped rapidly. And it was fragmented to all domains that were previously not used. And you can see the very messy stuff here is when the traffic was fragmented. But eventually, after a couple months, the traffic returned back to the primary domains.

So there are a lot of other things in the paper. We built a network of forum users, and we found that they were actually more connected after the tech (?) that happened. We also discuss the details of industry response and also the reactions of the leading community, where we found that they lost their interest quickly over time - just after a few weeks. And also the user discussion about the campaign tailed off very quickly.

So this very special case teaches us some valuable lessons. Even with a concerted effort by a series of tech firms - many tech firms - it may not be enough to shut down that hate community. And it can be quite challenging to do so, because one thing is missing. Because the Kiwifarms owner was not arrested. For example, when the FBI takes down a website, the playbook is that they seize the website, they arrest the owners, and they hope the combination may work. But in this case, the forum operator was not arrested, and he was very technically competent. He tried many things to recover the forum.

And we know from previous work that government actions may be less effective than a tech firm's, because they need to go through all the complicated paperwork, and they need to go through all legal processes before seizing any website. So the open question here is: How effective law enforcement and government action could be in that case? For example.


Yeah, I would like to spend my last word to say thank you to Ross. Because yeah, he was the one who changed my life in many ways. Thank you, Ross, for all the inspirations - all the things he has done for the community and for the generations and generations of students. And I have made a website. I collect all the public tribute of people that write about Ross - with consent of course - and I also collect some private messages by our department members, and I put them online. Because I believe that many people - including myself - will learn a lot of lessons by reading through people's memories. Thank you very much,

*clapping*

I thought this part was interesting:
So this very special case teaches us some valuable lessons. Even with a concerted effort by a series of tech firms - many tech firms - it may not be enough to shut down that hate community. And it can be quite challenging to do so, because one thing is missing. Because the Kiwifarms owner was not arrested. For example, when the FBI takes down a website, the playbook is that they seize the website, they arrest the owners, and they hope the combination may work. But in this case, the forum operator was not arrested, and he was very technically competent. He tried many things to recover the forum.

And we know from previous work that government actions may be less effective than a tech firm's, because they need to go through all the complicated paperwork, and they need to go through all legal processes before seizing any website. So the open question here is: How effective law enforcement and government action could be in that case?
Combine it with this slide:
23.png
Deplatforming an active community to suppress online hate and harassment without incapacitating the owners can be challenging!
empowering a regulator (Ofcom) to stop "harmful but not illegal" services may not be very effective.
I fully expected Anh V. Vu to strongly support increased regulations (he's a Vietnamese-British academic, after all), but he sounds pessimistic here. He seems to think that arresting or "incapacitating" @Null is the only thing that will actually be effective.
 

Attachments

Anh V. Vu said:
Here is the number of postings on Lolcow.farm, which is the primary competitor of Kiwifarms. So we do see some drop here in sync with Kiwifarms, but what interested me - users did not move to that forum. So they tend to move within their ecosystem.
Of course he's eyeing lolcow.farm. Once they take down KF they'll go after LCF and "competitor" sites like Ovarit too, despite the fact that they disallow doxing and are much more left-wing.

These trannies hate women having their own spaces to talk about rapist men in dresses on the Internet and will stop at nothing to dismantle them.
 
I went looking for a possible recording of Liz Fong-Jones' presentation at the University of Cambridge in November 2023.
Unfortunately, I think Liz probably requested that they not record it. If a recording existed, it should have / would have been archived here: https://www.cl.cam.ac.uk/research/security/seminars/archive/video/
I also found this: https://talks.cam.ac.uk/talk/index/207463
canceled-deleted.png
Liz definitely did the presentation (title: "Reverse engineering hate" - about his feud with Kiwifarms).
I forgot to point out that Liz was invited to come present at Cambridge by Ross Anderson.
From the screenshot above:
This talk is part of the Computer Laboratory Security Seminar series.
security-research-group.png
https://web.archive.org/web/20231128162202/https://www.cl.cam.ac.uk/research/security/people/
security-group.png
Before his death, Ross Anderson was the main guy in charge of Cambridge's "Security Seminars" where Liz presented. Alice Hutchings is listed there too. As a reminder: Ross and Alice were the co-authors for Anh V. Vu's academic paper about Kiwifarms.
co-authors.png
Just wanted to make the connection clear. And to make it even more explicit, Ross Anderson was actually the author of this blogpost:
https://www.lightbluetouchpaper.org/2023/11/09/how-hate-sites-evade-the-censor/
ross-anderson.png



If anyone wants some laughs, Anh V. Vu's paper has already been cited by five other papers:
https://scholar.google.co.jp/scholar?oi=bibs&hl=en&cites=8093741754001392850&as_sdt=5
I think the "Stoking the Flames" one is the funniest, and it's also specifically about KF. (The other papers are more broad.)
https://dl.acm.org/doi/pdf/10.1145/3641015
Online harassment remains a prevalent problem for internet users. Its impact is made orders of magnitude worse when multiple harassers coordinate to conduct networked attacks. This paper presents an analysis of 231 threads in Kiwi Farms, a notorious online harassment community. We find that networked online harassment campaigns consists of three phases: target introduction, network decision, and network response. The first stage consists of the initial narrative elements, that are approved or not in stage two and expanded in stage three. Narrative building is a common element of all three stages. The network plays a key role in narrative building, adding elements to the narrative in at least 80% of the threads, resulting in sustained harassment. This finding is central to our model of Continuous Narrative Escalation (CNE), that has two parts: (1) narrative continuation, the action of repeatedly adding new information to the existing narrative and (2) escalation, the aggravation of harassment that occurs as a consequence. In addition, we present insights from our analysis of 100 takedown requests threads, discussing received abuse reports. We find that these takedown requests are misused by the community and are used as elements to further fuel the narrative. We use our findings and framework to come up with a set of recommendations, that can inform harassment interventions and make online spaces safer.
9 CONCLUSIONS
In this work, we investigated the progression of online harassment attacks in Kiwi Farms, a hate and harassment forum. We present a three stage framework that explains how online harassment campaigns unfold. Our findings indicate that the narrative building starts with the initial post by the original thread creator and is further expanded by the network. Each narrative expansion, escalates the harassment directed to the target. Combining these two insights, we propose our model of Continuous Narrative Escalation (CNE). In the 231 threads we studied, the network contributes to the narrative with at least one element in 80% of the cases, indicating that the network plays an essential role in CNE. Further, we find that threads with a complete narrative are more likely to be sustained. We also uncover some of the strategies used in narrative escalation; continuous surveillance of targets’ online activity, misuse of abuse reports and any other evidence of target’s identity, content or activity. Ultimately we hope that our model can not only contribute to a deeper understanding of sustained harassment campaigns but also inform responses by the platforms, users, and the wider internet community. Towards this goal, we discuss the implications of our findings, including recommendations for enhancing safety of online spaces, informing interventions and designing better abuse reporting systems.

This section really jumped out at me:
8.2 Enhancing Safety of Online Spaces
In the past, threat intelligence stemming from online resources has also led to insights that helped combat human trafficking [47]. In the case of Kiwi Farms and other harassment forums, extracting threat intelligence insights might help anti-harassment efforts to identify emerging types and techniques of online harassment. In addition, our CNE model can be incorporated in automated abuse detection systems in online social networks that delegate moderation, such as Reddit where subforums, or "subreddits", are moderated separately from the site as a whole. The automated system may be used by the site’s administrators to scan various subreddits, detect the harassment and put a stop to it. Automated systems may also be used by federated social networks such as Mastodon to identify and defederate from instances that contain online harassment. Based on our model, natural language classifiers that detect the escalation of harassment (e.g., a harassment thread going through cycles of activity and inactivity) can complement current automated solutions that only look for explicit incitement and harassment language. This addition could also help enable de-escalation of the situation, for example by disabling the comment section or banning users involved in the escalation. Our findings also underline the need for multi-platform efforts to fight coordinated online harassment. The prevalence of content (in the form of URLs to other sites) in narrative building indicates that while harassers organize in a forum like Kiwi Farms, they may have identified ways to directly reach out to the target. While we are unable to measure if they actually do, we invite social networking sites to collaborate with external researchers to measure the extent of such cross-platform coordination, as well as work towards monitoring these communities for abuse that may be targeting the platforms themselves. One method of intervention could be based on observing a list of sites and communities known to proliferate harassment. Threat intelligence extracted from these sites could be used in generating signals informing nudges or warnings. These interventions would be notifying a creator to adjust privacy settings and mount a defense when their profile is being used in an attack.
AI moderation on social media sites is going to become more & more draconian. The specific mention of Mastodon brings to mind Liz Fong-Jones' interest in & frequent reposts of OpenTelemetry news on LinkedIn. I remember one of them was about OpenTelemetry being integrated into Mastodon.
Who wants to bet that Liz's obsession with "observability" and telemetry is at least partially because of his burning desire to quickly identify & shut down anything he considers to be hate/harassment/etc? And that would include anything about his "Consent Accident" too, of course! :)

I'll attach the full paper as a pdf to the bottom of this post, btw. I recommend reading it.
Oh, and check this out:
10 ACKNOWLEDGMENTS
This manuscript benefited from excellent feedback from anonymous reviewers and associate chairs at ACM CSCW. We would also like to thank fellow Ph.D. students and friends who helped proofread this paper. This work was supported by the National Science Foundation (grant number 2016061). Victoria Zhong was supported by a GAANN Fellowship awarded by the US Department of Education. Our lab has also received gifts from Google.
Google, the National Science Foundation, and the US Department of Education funded this. And the paper was published in Proceedings of the ACM on Human-Computer Interaction. The Association for Computing Machinery (ACM) is pretty prestigious.
 

Attachments

Who wants to bet that Liz's obsession with "observability" and telemetry is at least partially because of his burning desire to quickly identify & shut down anything he considers to be hate/harassment/etc? And that would include anything about his "Consent Accident" too, of course! :)
They're both buzzwords for tracking how well an application or suite of applications performing. What you're seeing in those posts is Elliot paying the bills as he'll often get to mention Honeycomb (the product that nobody uses or wants) in the same breath.
 
Who wants to bet that Liz's obsession with "observability" and telemetry is at least partially because of his burning desire to quickly identify & shut down anything he considers to be hate/harassment/etc?
I mean maybe? But that's literally all SREs do all day. Get new data, make shitty graphs and alerts, be fucking useless when anything bad actually happens.
 
They're both buzzwords for tracking how well an application or suite of applications performing.
And as buzzwords, they're meant to make you think there's something unique about something that's been being done since software even existed.
I fully expected Anh V. Vu to strongly support increased regulations (he's a Vietnamese-British academic, after all), but he sounds pessimistic here. He seems to think that arresting or "incapacitating" @Null is the only thing that will actually be effective.
Pretty mask off, admitting he wants to murder people for disagreeing with his totalitarian ideology. But yeah, we're totally the bad guys, not these criminal freaks hiding behind academia to ruin and destroy society and basic principles of freedom.

These other two should join Ross Andrews in Hell as soon as is convenient for them.
 
Back