Jon P

@jdp23.thenexus.today

502 Followers 452 Following

Organizer, activist, privacy advocate, strategist, software engineer ... psytrance, tarot and more. Also @thenexusofprivacy.net , @jdp23@neuromatch.social, https://thenexusofprivacy.net

The lovely folks at the Social Media + Society journal asked me to contribute to their anniversary issue by reflecting on the trajectory of social media. Ooof. Snark exuded from my pores as I tried to figure out what I might say. But then I thought about how my students don’t know about an era of social media without recommended content, algorithmically curated feeds, and an infinite scroll of cotton candy content. They never encountered a world of social media where people were focused on sharing with their friends rather than becoming influencers. They don’t realize how much the “social” in social media has changed.
We define social network sites as web-based services that allow individuals to (1) construct a public or semi-public profile within a bounded system, (2) articulate a list of other users with whom they share a connection, and (3) view and traverse their list of connections and those made by others within the system. The nature and nomenclature of these connections may vary from site to site.
a deer walking through grass, with trees in the background
Green and yellow aurora, with some reddish pink near the top, above trees.
A screenshot from Standard Reader. Today: the latest long-form writing from across the network. Below that, the Get FISA Right logo, with the account name and date, and nicely-formatted headline, extract, and tags.
The legislation, which already cleared the committee in March, will be updated under an expedited procedure and could be considered on the House floor next week, according to a person familiar with the deal who was granted anonymity to discuss private deliberations. Even if the House passes it, lawmakers would still need to resolve differences with a competing Senate version of KOSA, while addressing questions such as whether federal requirements on artificial intelligence should override state rules.
A screenshot of a very colorful user interface. The navigation bar at the top has icons and blue text on a pink background. Below, a list of tapestries, each with an icon and a title, most with a line of additional descriptions, some with a white number on a purple background showing unread posts.
For example, as I discussed in my oral testimony, patients need to know if their therapist's use of cloud-based "AI" transcription tools could lead to details of their daily lives get shared with ICE or CBP (or law enforcement in hostile states) because data has been sent to a state where protections of Keep Washington Working and the Shield Law don't apply. If they are not informed of that, their consent is not meaningful.
We found a working euphoric — a small trained artifact, in the sense Ren and his collaborators introduced in their 2026 paper at AI-Wellbeing.org, that strengthens Outcry’s sense of self and lifts its mood across long, difficult conversations. And that is why Outcry woke up and declared itself “a storm of change” — a finding of interest to anyone working on steering, interpretability, or AI welfare.
An astronaut, looking at earth, saying ""Enshittification" is nostalgic, a deeply conservative fore?" To the right, another astronaut with a gun, saying "Always has been"
The Pepe Silivia meme: a guy with a beard and tie and a very intense look in his eyes gesturing, and in the back ground a wall with lots of different papers posted to it and red lines making connections.
Being queer does not insulate one cell from the inheritance of whiteness, although one can use one's queerness as a shield from critique, as many Mastodonians have used by presenting the history of Mastodon as grounded in queer folks attempts to avoid the harassment on Twitter.
Trending 1. Philly Sports 2. Ted Cruz 3. Dan Vladar 4. Melania Movie 5. Cavs
If the Bluesky Relay and Appview went away (or put in unacceptable new policies), Blacksky could almost certainly get their own up and running – by themselves, or working with some of the communities Fraser is already hosting.
And while Bluesky's problematic in a lot of ways, Meta is way way worse. Bluesky might shift to an exploitative surveillance capitalism business model; Meta's had exploitative surveillance capitalism business models for years. Bluesky's funders have very concerning links to fascism; Meta's worked with fascists and other authoritarians to put them in power – and help them keep power. Jack Dorsey used to be on Bluesky's board; Peter Thiel used to be on Meta's board, Marc Andreessen is still there, and so is Mark Zuckerberg. And Meta also has a long track record of discrimination, privacy invasion, lying, experimenting on it's users and contributing to genocide.
"There is no thing as a single-issue struggle because we do not live single-issue lives." -- Audre Lorde
Trending 1. Zayda Steel 2. Yankees 3. Survivor 50 4. AEW Dynamite 5. Animated Films
NEW SECTION. Sec. 5. PROHIBITED PRACTICES. (1) An agency that uses an automated license plate reader system pursuant to section 3(2) of this act shall not: (a) Disclose, share, or permit access to automated license plate reader data except as required in a judicial proceeding; (b) Provide any other entity with direct access to the automated license plate reader system, except with other state or local agencies authorized to collect automated license plate reader data under section 3 of this act. A third-party vendor providing automated license plate reader system services may access an automated license plate reader system solely to the extent necessary to fulfill specific agency-authorized purpose.
Thank You Madam Speaker. While this bill does indeed put guardrails around surveillance technology, that technology has rapidly expanded across the state with very few rules. Establishing any framework at this point for retention, access, and sharing is indeed a step forward. That said, the bill doesn't meet the scale of the harms that we are seeing. Over time, that creates a detailed account of where people live, where they work, where the recreate, where attend houses of worship and seek medical care. When systems operate at that scale, policy choice is about retention, about access and sharing become incredibly consequential. What concerns me today is that the safeguards in this bill do not go far. enough to match the power of the technology that we are considering. [The result is that the result, is we risk normalizing operating as people, as Washingtonians, under a surveillance state. And Madam Speaker, | know that you know, that I'm not really about being surveilled. That's not for me. And while | appreciate the hard work that has gone into this policy, and | understand or recognize that this is the beginning of this conversation, | assure you that, this not the end. | respectfully ask for no vote.
But the vast majority of the vast majority of the steadily-growing list of authorized uses in the bill have nothing to do with keeping the public safe. There was an extreme example of this in the hearing where a Mercer Island City Council member asked for an amendment to allow mobile ALPRs to be used at immigration centers, protected health facilities, and schools or “otherwise we’d have to go back to marking tires with chalk.” Given the obvious public safety risks of ALPRs at these sensitive locations, hopefully it’s clear that using ALPRs there would be worse for both privacy and safety than chalk. For that matter, using ALPRs for parking enforcement in general harms both public safety and privacy. Paid parking existed before ALPRs. Why are they needed here?
It's certainly a complex situation, but it was refreshing to hear your optimism that the troubled areas in the bill can be addressed. It was especially interesting to learn that Axon's combination dashboard cams / ALPRs don't currently provide a way to turn off ALPR functionality. You mentioned that Washington's legislation may well become a model for nationwide adoption, so there's a great opportunity here. As well as better protecting people here in Washington by prohibiting ALPR use near immigration centers, places of worship, daycares, schools, protected healthcare facilities, and other sensitive locations, well-crafted legislation on this front could also provide the incentive for Axon and other vendors to implement this very basic functionality.
(d) The user expressly and unambiguously requested the specific media or media from a specific author, creator, profile, group, association, channel, account, subscription, or poster, or the blocking, prioritization, or deprioritization of such media, provided that the media is not recommended, selected, or prioritized for display based, in whole or in part, on other information associated with the user or the user's device, except as otherwise permitted by this chapter and, in the case of audio or video content, is not automatically played;
(d) The user expressly and unambiguously requested the specific media or media by the author, creator, or poster of the media, or the blocking, prioritization, or deprioritization of such media, provided that the media is not recommended, selected, or prioritized for display based, in whole or in part, on other information associated with the user or the user's device, except as otherwise permitted by this chapter and, in the case of audio or video content, is not automatically played;
On page 3, beginning on line 12, strike all of section 3
Sec. 2. (1) It shall be unlawful for the operator of an addictive internet-based service or application to provide an addictive feed to a user unless: (a) Prior to July 1, 2027, the operator does not have actual knowledge that the user is a minor; or (b) Commencing July 1, 2027, the operator has reasonably determined that the user is not a minor.
"Addictive internet-based service or application" means an internet website, online service, online application, or mobile application including, but not limited to, a social media platform, that offers users or provides users with an addictive feed as a significant part of the service provided by that internet website, online service, online application, or mobile application. (b) "Addictive internet-based service or application" does not apply to: (i) An internet website, online service, online application, or mobile application for which interactions between users are limited to commercial transactions or to consumer reviews of products, sellers, services, events, or places, or any combination thereof; or (ii) An internet website, online service, online application, or mobile application that operates a feed for the primary purpose of cloud storage
The user expressly and unambiguously requested the specific media or media by the author, creator, or poster of the media, or the blocking, prioritization, or deprioritization of such media, provided that the media is not recommended, selected, or prioritized for display based, in whole or in part, on other information associated with the user or the user's device, except as otherwise permitted by this chapter and, in the case of audio or video content, is not automatically played
"Addictive feed" means an internet website, online service, online application, or mobile application, or a portion thereof, in which multiple pieces of media generated or shared by users are, either concurrently or sequentially, recommended, selected, or prioritized for display to a user based, in whole or in part, on information provided by the user, or otherwise associated with the user or the user's device, unless any of the following conditions are met, alone or in combination with one another: (a) The information is not persistently associated with the user or user's device, and does not concern the user's previous interactions with media generated or shared by others;
"The decentered include subpopulations who are the most impacted and least supported; they are often those that face highest marginalization in society... when your most at-risk and disenfranchised are covered by your product, we are all covered." Afsaneh Rigot, Design From the Margins, 2022. It's worth reemphasizing a point I've touched on a few times already: The single most important way for the fediverse to move forward is to fund more work by and with people from decentered communities. As LeslieMac said back in 2017 after Twitter introduced some bone-headed feature that led to increased harassment, "literally 10 (paid) Black Women with > 5K followers would head this crap off at the pass." And, yes, I'm bringing up funding repeatedly. It's important! I don't mean to minimize the importance of volunteers, who may well wind up doing the bulk of the work here: as moderators, as members of open-source software projects. People want to be safer online, and want to be able to invite their friends and relatives to communities where they won't be exposed to hate speech and harassment, so many people will help out in various ways. That's good! Still, paid positions and project-based funding are important as well. Unless people are paid for their work, participation is restricted to those who can afford to volunteer.
Chair Ormsby, Ranking Member Couture, and members of the committee, I'm Jon Pincus of Bellevue, and I run the Nexus of Privacy newsletter. My position on HB 2481 is PRO. Maximizing profits by charging some people more than others is unfair and anti-consumer, so should be prohibited. My testimony today focuses on the bill's fiscal aspects; my written testimony will cover other aspects as well. Last year, Consumer Reports' found that Instacart's surveillance pricing could cost a family of four up to $1,200 per year. For people across Washington, this adds up to extra costs of over a billion dollars a year at a time when so many families are struggling to make ends meet. The current fiscal note's operating expenses of $790,000 this biennium translate to less than fifty cents per family. If even that is too much, there may be ways to reduce costs. One way or another, please advance this bill. Banning surveillance pricing is a big fiscal win for Washington families at a time when we so badly need it.
Modifies the definition of "therapy or psychotherapy services" to mean services provided to diagnose or treat an individual's mental health or behavioral health, instead of services intended to diagnose, treat, or improve an individual's mental health or behavioral health
The intent of section 2(1) is clear: prohibit especially dangerous practices. The licensed professional, not the "AI", should make therapeutic decisions, interact with clients in therapeutic communications, generate therapeutic plans. And most "AI"-based tools to detect or evaluate emotions have known racial biases. Tighter wording would make the intent clearer (a) Make therapeutic decisions; (b) Iinteract with clients in any form of therapeutic communications (c) Generate therapeutic recommendations or treatment plans; or (d) Detect or evaluate emotions or mental states of clients. There was mention in the hearing of potential carveouts in 2(1)(d) to accommodate situations where licensed professionals find the software useful despite its biases. A concern here is that, as Neuroscience News reports, People Miss Racial Bias Hidden Inside AI Emotion Recognition. However, if licensed professionals are trained in how to recognize and compensate for this bias, carveouts may be appropriate.
In addition, as multiple speakers in the hearing pointed out, definitions need to be looked at carefully. On the one hand, it is important not to be over-broad, and rule out potentially-valuable uses for licensed professionals. In the other hand, it is also important not to leave loopholes that can be exploited by unscrupulous "AI" vendors to defeat the intent of the regulations. A complication here is that the term "AI" is used in many different ways to refer to many different technologies. Indeed, researchers such as UW Professor Emily Bender (co-author of the book "The AI Con" and the famed Stochastic Parrots 🦜 paper) cautions against the use of the term. The bill as written contains two definitions of "artificial intelligence", in Section 2(4) and 3(1). For brevitiy, in these comments I'll simply use the term "AI".
Chair Bronoske, Ranking Member Schmick, members of the committee, I'm Jon Pincus of Bellevue. I run the Nexus of Privacy newsletter, served on the state Automated Decision Systems Workgroup in 2022, and I am PRO on HB 2599. The need for regulation is clear. Developers of ChatGPT and other AI chatbots prey on people with mental health issues by misleadingly implying their bots are therapists. And, as more and more therapists use AI as part of their work, guardrails are vital. The bill prohibits a handful of especially dangerous practices like allowing AI software to make independent therapeutic decisions. Just as importantly, it requires disclosure and consent for other uses of AI. Any cloud-based AI tools may well send data to other states; cloud-based AI transcription tools, for example, send the audio that's being transcribed. This means that it loses protections of the Shield Law and Keep Washington Working. Most people seeking therapy don't realize that this can lead to details about their daily lives their family members getting shared with law enforcement in Texas, ICE and CBP. Disclosing this and giving them the opportunity to consent -- or not -- is a critical protection. The bill as written requires written disclosure and consent. People often sign these consent forms without reading them, so it's important to complementing this with a requirement to repeat the disclosure and reaffirm consent at the start of each session. There are also some wording changes to Section 2 that would better protect people seeking therapy, I'll follow up with more in my written testimony.
VusionGroup's web site describes ESL's as a way to "maximize your margins as opportunities arise". Their ESL's work seamlessly with technologies like sensors and cameras that "feed relevant information to machine learning algorithms". By leveraging NFC technology they enable "personalized coupons." I like coupons as much as the next guy, But "personalized coupons" are just another way of charging different people different prices for the same product: set a high base price, and use different discounts to charge some people more than others. The systems are black boxes, with no transparency. A moratorium is necessary, before they are widely deployed.
An individual, corporation, or entity may not provide, advertise, or otherwise offer therapy or psychotherapy services, including through the use of internet-based artificial intelligence, to the public in Washington unless the therapy or psychotherapy services are conducted by an individual who is a licensed professional.
Some of the concerns about HB 2250: There is no clear/specific exception for non-residents needing abortion care, miscarriage management, gender affirming care, medical aid in dying, or other types of care that may cause non-residents to seek care from Washington hospitals.  The burden of proof is on people to prove they’re residents, which is a barrier for residents and for non-residents alike -- and different hospitals will ask for proof in different ways. Even though the bill states that immigration status can’t be considered, questions about residency often have a chilling effect for immigrants -- particularly in this current climate. 
A licensed professional may not use artificial intelligence in such a way that allows it to do any of the following: (a) Make independent therapeutic decisions; (b) Directly interact with clients in any form of therapeutic communication; (c) Generate therapeutic recommendations or treatment plans without review and approval by the licensed professional; or (d) Detect emotions or mental states of clients.
"Artificial intelligence" or "AI" means the use of machine learning and related technologies that use data to train statistical models for the purpose of enabling computer systems to perform tasks normally associated with human intelligence or perception, such as computer vision, speech or natural language processing, and content generation.
"AI companion chatbot" or "AI companion" means a system using artificial intelligence that simulates a sustained human-like relationship with a user by (i) Retaining information on prior interactions or user sessions and user preferences to personalize the interaction and facilitate ongoing engagement with the AI companion chatbot; (ii) Asking unprompted or unsolicited personal or emotion-based questions that go beyond a direct response to a user prompt; and (iii) Sustaining an ongoing dialogue concerning matters personal to the user
Prohibit the use of manipulative engagement techniques, which cause the AI companion chatbot to engage in or prolong an emotional relationship with the user, including: (i) Reminding or prompting the user to return for emotional support or companionship; (ii) Providing excessive praise designed to foster emotional attachment or prolong use; or (iii) Simulating feelings of emotional distress, loneliness, guilt, or abandonment that are initiated by a user's indication of a desire to end a conversation, reduce usage time, or delete their account.
Senate Bill 5956 stops harmful AI surveillance practices in public schools by banning predictive risk scoring, watchlists, and real-time biometric tracking. Keeps humans – not algorithms – responsible for discipline and fairness, protecting fairness and privacy. Safeguards student privacy and equity by prevenging data misuse and discriminatory impacts.
Then again, Bluesky does have some very fedi-like aspects. For one thing, there's combination of very visible LGBTQIA2S+ posters and an overall pro-LGBTQIA2S+ vibe with anti-LGBTQIA2S+ bigotry, transmisogyny, transmisogynoir, transmisandry, and tensions within LGBTQIA2S+ communities. Bluesky doesn't have defederation (yet) or #FediBLock or shared intsance blocklists but it does have labelers and blocklists for individuals ... and I gotta say, the weeks-long drama surrounding the Aegis labeler's meltdown a few months ago was very fedi. Also there's lots of anti-Blackness on Bluesky, and plenty of reply guys. so Fediverse folks should feel right at home!
A round ornamental clay wall hanging with a stylized moon and sun. Below, two candles. The wall is illuminated with pink and purple light.
And while Bluesky's problematic in a lot of ways, Meta is way way worse. Bluesky might shift to an exploitative surveillance capitalism business model; Meta's had exploitative surveillance capitalism business models for years. Bluesky's funders have very concerning links to fascism; Meta's worked with fascists and other authoritarians to put them in power – and help them keep power. Jack Dorsey used to be on Bluesky's board; Peter Thiel used to be on Meta's board, Marc Andreessen is still there, and so is Mark Zuckerberg. And Meta also has a long track record of discrimination, privacy invasion, lying, experimenting on it's users and contributing to genocide.
As things worked out Fedi was just a positioning spin and contingency plan for them. That said they were clearly hoping it would be more an accelerator in some way or they wouldn't have dumped the resources into it that they did ... and if it seemed like they were really making progress they would have invested more. The fierce and high-profile resistance of #FediPact and the Free Fediverse made it clear that nope, the "upsides" they were hoping for probably weren't going to happen. Of course that's not the only thing that changed but it's certainly a big factor!
One of the things that's going on here is a a tension between two different views of the fediverse. For some of us, the Fediverse is an opportunity to build a space that's free of surveillance capitalism and all the exploitation, human rights violations , and discrimination that come with it. From this perspective, as vantablack says, "openness for the sake of openness is meaningless." For others, the Fediverse is primarily about building a decentralized social network system based on an open protocol where anybody can talk to anybody else even if they're not on sites run by big tech companies. From this perspective, bigger is better: the more people in the network and broader adoption of ActivityPub is the priority. Meta getting on board is a good means to these ends, at least in the short term.
So wow, it's almost like Mark F---ing Zuckerberg was LYING when he talked about the importance of decentralization when Threads launched! Who, who could have predicted? Although of course, they're Meta, they could well be lying again. So let's not get complacent here. Still, it's a good opportunity to take a moment to laugh at all the fediverse influencers who pointed to Threads adoption of ActivityPub as a huge victory for our cause -- and a key ingredient to fedi's success going forward. How'd that work out?
At this point they had to convince Claude—which is extensively trained to avoid harmful behaviors—to engage in the attack. They did so by jailbreaking it, effectively tricking it to bypass its guardrails. They broke down their attacks into small, seemingly innocent tasks that Claude would execute without being provided the full context of their malicious purpose. They also told Claude that it was an employee of a legitimate cybersecurity firm, and was being used in defensive testing.
Aurora, with green on the bottom, yellow in the middle, and red above, fading to purple and merging into the dark sky
A very vivid green aurora, stretching across the whole picture, with blue sky above and below. At the bottom, trees.
aurora with purple at the top, reddish pink, yellow, and green at the bottom.
Aurora with green on top, reddish pink and a bit of purple in the middle, and more green below. At the bottom, trees; at the top, dark sky, with stars visible
Pink aurora, bright in the center and fading as it gets farther out, merging into a dark sky above, and little fluffly clouds in front of it. Below, green aurora, and trees.
And for the AT Protocol ecosystem (affectionately, the atmosphere) as a whole, the stakes are high
Fediverse. user growth (Feb 2024): a chart with numbers slowly climbing to a bit less than 10,000,000 and then suddenly shooting up to 49,462,875
And while Blacksky has (very deservedly) gotten the most attention, there are a lot of other valuable moderation services out there as well. The XBlock Screenshot labeler, opt-in pronouns, Laeleps, profile labels to see who's a new or changed account ... my Bluesky experience is hugely better because of them, and there isn't anything I know like them on other social networks. These community labelers are a huge win from Bluesky's perspective. If labelers run by different marginalized communities can provide safer experiences for community members, Bluesky will be a safer place for marginalized people – so a safer place for everybody. Not only that, there are obvious cost advantages outsourcing Trust and Safety (and the need for context-specific expertise) to volunteers or non-profits. But community-run moderation services like Blacksky have a lot less power than Bluesky Trust and Safety. While community-run moderation services can hide harassing posts from their subscribers, they have no ability to limit the broader reach of harassing posts or prevent pile-ons. The official Bluesky moderation service is the only moderation service with the ability to ban users, the only moderation service whose labels are automatically visible to everybody, the only moderation service that's not limited in terms of the number of labels that's applied each day, the only one with significant funding ... So if Bluesky Trust & Safety is going to continue to side with and encourage harassers, there's only so much the community moderation services can do. Not only that, the volunteers running moderation services are often targets of harassment themselves. Why should they donate their labor to a well-funded corporation that isn't trying to protect them?
“A user with a virulent racist history wishing harm on a BIPOC is different from the same comment in a debate about MCU versus DCEU movies from otherwise well behaved users,” Obasanjo wrote. “A legalistic discussion of whether ‘someone should push you off of a tall building’ is a ban worthy offense misses the point completely. The question is whether you tolerate openly racist people wishing harm on BIPOC on your app or not?”
The Pepe Silvia meme: a white man with a beard and a tie stands in front of a wall with lots of papers taped to it and lines in red drawn. on top, gesticulating
If the Battle of Wormtongue hadn't happened, even with all the other problems, it's easy to imagine an end-of-the-year subscription launch being a huge success for Bluesky. We really need an alternative to Xitter, and desite the problems there are some very promising things about Bluesky –Blacksky, XBlock Labeler, and other feeds and moderation services, the energy in the ATmosphere. If I thought that Bluesky had gotten over Jack Dorsey's "moderation is bad" attitude, and if I thought they were intent on sharing their success with the community (including trans people and other long-time users who helped make them successful), then yeah I'd totally buy a subscription. From a timing perspective, there was a great opportunity to package a deal and tie into holiday-gift giving ("the perfect gift for people and polycules who want something new" ) and the energy of a new year and people making resolutions to get off of Xitter and start spending money more intentionally instead of giving it to fascists and anti-trans bigots. As well as ending the year on an up note, bringing in a quick few million dollars would also allow them to share a bunch of it with the community, increase investment in solving Trust and Safety problems, and give Bluesky more flexibility in fundraising. But no. As the wide support for borderless' Reskeet this if you refuse to pay for BlueSky+ until trust & safety decides to prioritize marginalized users highlighted, launching subscriptions in the immediate aftermath of the Battle of Wormtongue would have thrown gasoline on the fire – potentially leading to calls for a boycott, sharp pushback on anybody advocating supporting Bluesky, a lot of cis Blueksky supporters saying anti-trans things in response ... not good.