plant lover, cookie monster, shoe fiend
19691 stories
·
21 followers

How ChatGPT Could Be Making Your OCD Worse | Teen Vogue

1 Share

From time to time, we all have questions that boil down to Is this normal? Did I do the right thing? Am I okay? About two years ago, Kate — who asked to use only her first name for her privacy — started typing these kinds of questions into ChatGPT.

“Nobody has a guide for being human that shows you a manual of all the ways that are normal to act,” she said. “I guess it’s like [I was] looking for that authoritative source that goes, ‘Yes, this was certainly the right way or the wrong way or the abnormal way to act.’” Feeding it a scenario from her life, she’d ask whether she could have misinterpreted something or if she did the right thing. “It doesn’t really answer the question, because nobody can answer the question,” she added.

Even though Kate knew she couldn’t get the certainty she wanted, she would sometimes spend up to 14 hours a day posing these kinds of questions to ChatGPT. “You want it to reaffirm, to add weight,” she said. “If you're 99% sure, you want it to make that 100, but it can't because that's not a thing.”

This urge to ask for assurance again and again can amount to compulsive reassurance-seeking, which is common among people with anxiety disorders and obsessive-compulsive disorder. We all need some affirmation on occasion, but what makes compulsive reassurance-seeking different is that someone will linger on a bit of doubt trying to reach nonexistent certainty, according to Andrea Kulberg, a licensed psychologist who has been treating anxiety for 25 years.

“People do it because it gives them the illusion of certainty,” Kulberg said. By researching online or asking questions to a chatbot, you’re trying to convince yourself that something bad won’t happen, she explained. And while securing reassurance may offer a temporary bit of relief, it actually gives credence to the need to seek reassurance and can increase anxiety over time. Kulberg added, “The anxiety never says, ‘We're good, okay, you can stop reassurance seeking,’ because it's always followed by more doubt.”

There are many avenues people use to compulsively seek reassurance — books, forums, Google, friends and family. But unlike AI chatbots, these other resources don’t prompt their users to keep going, which is one of the features that can make AI chatbots a perfect storm for individuals with OCD and anxiety disorders. “It never gives you one complete response,” Kate said. “It always says, ‘Would you like me to do this?’ And I'm like, well, yeah, sure, if we're not finished, if it's not complete.”

“It’s a massive wormhole for me,” said Shannon, who can spend upwards of 10 hours a day asking for reassurance from AI chatbots. (Shannon also asked to use only her first name.) She keeps several chats active, each reserved for a particular topic that her anxiety regularly hones in on. “I'm definitely aware that it's not healthy to do. I do try to avoid it, but I still find myself getting sucked in,” she said. “I'll just think of something, and I'll just feel that urge to go and ask AI about it.”

Occasionally, when Kate questions ChatGPT for hours about a single topic, the chatbot eventually tells her there is nothing else it can say on the matter. “I think most people never get to that point where it goes, ‘I give up,’” she said. But other than these moments — or when her phone battery dies — there are few breaks.

Read the whole story
sarcozona
45 minutes ago
reply
Epiphyte City
Share this story
Delete

Secret changes to major U.S. health datasets raise alarms

1 Comment

A new study in the medical journal The Lancet reports that more than 100 United States government health datasets were altered this spring without any public notice. The investigation shows that nearly half of the files examined underwent wording changes while leaving the official change logs blank. The authors warn that hidden edits of this kind can ripple through public health research and erode confidence in federal data.

To reach these findings, the researchers started by downloading the online catalogues—known as harvest sources—that federal agencies maintain under the 2019 Open Government Data Act. They gathered every entry from the Centers for Disease Control and Prevention, the Department of Health and Human Services, and the Department of Veterans Affairs that showed a modification date between January 20 and March 25, 2025.

After removing duplicates and files that are refreshed at least monthly, the team was left with 232 datasets. For each one, they located an archived copy that pre‑dated the study window, most often through the Internet Archive’s Wayback Machine.

They then used the comparison feature in a word‑processing program to highlight every textual difference between the older and newer versions. Only wording was assessed; numeric tables were not rechecked. Finally, the investigators opened the public change log that sits at the bottom of each dataset’s web page to see whether the alteration had been declared.

One example captures how the edits appeared in practice. A file from the Department of Veterans Affairs that tracks the number of veterans using healthcare services in the 2021 fiscal year had sat untouched for more than two years. On March 5, 2025, the column heading “Gender” was replaced with “Sex.” The same swap was made in the dataset’s title and in the short description at the top of the page. The modification date on the site updated to reflect the change, yet the built‑in change log still reads, “No changes have been archived yet.”

Across the full sample, the pattern was strikingly consistent. One hundred fourteen of the 232 datasets—49 percent—contained what the authors judged to be potentially substantive wording changes. Of these, 106 switched the term “gender” to “sex.” Four files replaced the phrase “social determinants of health” with “non‑medical factors,” one exchanged “socio‑economic status” for “socio‑economic characteristics,” and a single clinical trial listing rewrote its title so that “gender diverse” became “include men and women.”

In 89 cases, the revision affected text that defines the data itself, such as column names or category labels. The remaining 25 changes occurred in narrative descriptions or tags that sit above the data table. Only 25 of the 114 altered files—less than one in seven—acknowledged the revision in their official logs.

The timing followed a marked acceleration: four edits occurred in the final days of January, 30 during February, and 82 during the first three and a half weeks of March—suggesting an intensified push as spring approached.

These government datasets form the backbone of countless psychology, sociology, and public health projects. The Behavioral Risk Factor Surveillance System, for instance, supplies yearly survey information on smoking, exercise, diet, and chronic illness across every state. It is routinely mined to study links between health behavior and mental well‑being.

Heart disease and stroke mortality files from the Centers for Disease Control and Prevention help social scientists examine how stress, neighborhood environment, or discrimination align with geographic patterns in illness and death.

Nutrition and physical activity surveys inform work on childhood obesity and its ties to screen time or family structure. Researchers who focus on veteran mental health rely on Department of Veterans Affairs summaries to track service‑connected disability, access to therapy, and suicide risk among former service members.

When variable labels shift from “gender” to “sex” in these resources, studies that compare answers given under the old wording with figures retrieved after the change are no longer aligning like‑with‑like. Even a single undocumented edit can scramble replication attempts, invalidate earlier statistical models, or make it impossible to detect real trends in the underlying population.

The implications stretch beyond statistical concerns. Survey designers distinguish between gender, a social identity, and sex, a biological classification, because the two terms capture related but not identical information. Many transgender and non‑binary respondents, for example, select a gender option that differs from the sex recorded on their birth certificate.

If the government retroactively re‑labels a column without clarifying whether the underlying question also changed, analysts cannot tell whether a fluctuation in the male‑to‑female ratio reflects genuine demographic shifts, a wording tweak, or recoding behind the scenes. Public health officials may then allocate resources on a faulty premise, and medical guidelines that depend on demographic baselines can drift off target.

The authors of the study point to a possible political origin for the edits. They note that the White House issued a directive in early February instructing agencies to purge material seen as advancing “gender ideology”—language echoed by several state administrations.

No federal office has publicly confirmed that the dataset edits were carried out in response, yet the timing and the tight focus on the term “gender” hint at coordinated action. If the goal was to bring terminology across agencies into alignment, the transparency required by the Open Government Data Act appears to have been set aside.

The investigation is not without limits. Because many archives extend back only a few years, the researchers could not examine earlier periods for similar actions. They judged whether a change was routine or substantive by hand, an approach that introduces subjectivity. They also left numerical content untouched; it remains unknown whether any figures were edited alongside the wording.

In response to the findings, the authors suggest a series of steps that scholars and institutions can take to protect the reliability of public data. Independent groups already mirror many federal datasets on private servers, and individual investigators can save local copies of files they intend to analyze. Routine spot checks against archived versions can help reveal unexpected alterations.

International repositories such as Europe PubMed Central offer alternative hosting for biomedical resources, lowering dependence on any single government. Most important, the researchers argue, is a cultural commitment to full version tracking inside federal agencies—so that every member of the public can see exactly what changed, when it changed, and why.

The study, “Data manipulation within the US Federal Government,” was authored by Janet Freilich and Aaron S. Kesselheim.

Read the whole story
sarcozona
50 minutes ago
reply
Oh no
Epiphyte City
Share this story
Delete

Categories Of Bad, Home, Coalition, And Theory of Change | Cogito, Ergo Sumana

1 Share

(This is one of the "people are wrong on the internet!!!" posts I write to clear it out of my head. And then refer back to later in future arguments, in imagined triumph, as though "Well I conclusively demonstrated that YOU are WRONG more than EIGHT YEARS AGO in a POST on my BLOG" is going to change anyone's mind. Well, they aren't all gems of persuasion and grace.)

Categories can be important, especially when it comes to behaviors we abhor. And I believe that when we allow ourselves to be sloppy about labelling harmful actions, we make it harder to make the right decisions about how to deal with them.

I'll lay this out in more detail.

How it comes up

I hang out in groups that aim to avoid bigotry. We don't want to be bigoted, and we want to protect ourselves and others from bigotry coming from outside. And we know that dominant cultures tend to brush off and minimize bigoted things people say and do: "He didn't mean it" and "It's not a big deal". Further, we know that it's human nature to tend to give our friends the benefit of the doubt: "Well I know it looks bad but you have to understand the context!"

So we counter that with a practice -- sometimes deliberate, sometimes reflexive -- to notice and explicitly speak about even subtle bias that crops up. We want to root out not only blatant bigotry, but also the small pernicious ways systemic bias comes out in our words and deeds.

We should be on the watch for bigotries! And we should fight back in a variety of ways! But sometimes, I'm in groups where the following dynamic happens:

Person A: This bad thing happened. I hereby claim that it was abusive, or violent, or otherwise so abhorrent that we ought to shun or otherwise punish the person who did it, and/or that we should no longer make use of their work.

Person B: I am not sure I agree with the claim that the bad action deserves that particular label. Or: I am not sure that the response you suggest is appropriate.

[Person A and/or others]: It is illegitimate for you to doubt the accusation or the suggested response. Sometimes: It is illegitimate for anyone to ever try to make fine distinctions about those definitions, and to do so is an aid to bigotry.

And of course it's usually messier than that, with heated emotions, people moving goalposts, and so on.

The range of possible responses

I wrote a little of this while talking with other MetaFilter users about moderating misogynistic content, in case you want to read an example of this dynamic in action. And in that context, I said:

But when we differ in our names for various categories of bigotry then it's harder for us to get on the same page and then say "here's how we should react to this category" (like, delete category A + ban user, delete category B + warn user, delete a category C post + email poster to ask whether they can find a different source that isn't gross, heavily mod category D and support users who put out a Bat-Signal to ask for in-thread participation from people affected by that particular oppression, etc.).

More generally: when we differ in our names for a particular kind of bigotry, or when we lump together different kinds of biased actions, then we have a harder time making good decisions about how to react. And this is especially important in a group context, where we have to make shared decisions about whether to complain, delete, exclude, and so on. One way to understand our range of possible responses, within a voluntary group, is that they range from "no response" up through "dilution" and "complaint" till we get to "removal".

  1. Removal: Shunning, blacklisting, and/or boycotting the person, and possibly their work as well. They can't come in, and we pre-emptively decide to ignore any or most input they offer. We decline to let them buy our products or services, and we decline to buy theirs, as an organized form of pressure aimed at getting them to apologize or to change a specific policy. (I decline to call something a boycott unless it's organized this way; if it's an expression of disgust without a goal of changing the other person's behavior, that can be a fine thing to do, but it's not a boycott.) And we "de-platform" them, which is to say that we don't share or publicize their work in ways that would possibly benefit them.
  2. Complaint: We criticize the person or work, possibly in public and/or possibly in private conversation with them, with suggestions for improvement. I think of this as reporting bugs. Note that it's pretty hard to simultaneously do this while shunning someone. This approach assumes that we're probably trying to get the person to change, and that we're therefore willing to work at dialogue -- clarifying, suggesting ways they could improve, and so on. Or maybe we can't or don't want to put the energy into that, but we still want to warn each other about how it might hurt us to get mixed up with them.
  3. Dilution: We dilute one person's negative impact by ensuring that they aren't the only source of information other people get. We create and publicize better information from several perspectives, and encourage others to do the same. If the problem is that they agree with us about our core issue but disagree with us about secondary issues, then we pre-empt their impact by proactively explaining that we're a big tent, and that it's okay and expected for lots of us to hold diverse perspectives on non-core issues.
  4. No response: We don't, as a group, change how we treat this person and their work.

(I'm riffing off a blog post I ran across several years ago, "Quality Control in Movements" by Burgess Laughlin, which suggests a taxonomy of available "tools for individuals who are trying to minimize the effect of counter-productive members of a movement". It's by an Objectivist who's willing to use phrases like "false representatives of the movement" but I still found it thought-provoking.)

We can make different choices, depending on our goals and resources, depending on what the person is doing and how it affects us, depending on whether we'd like to help them change. And we can escalate as new facts emerge. If we lump together everything from "still has an announcement-only Twitter account even though the platform now is much friendlier to hate speech" to "explicitly discriminates against and foments hate toward women," and respond to them all with complete removal instead of ever using complaint or dilution or even no response, then ....

See, here is where my argument depends on your perspective. Because there is a stark difference, in what I have to explain to you, depending on whether or not you have ever been responsible for any institution that materially affects other people, especially people you don't personally know. "Institution" here could mean a recurring party, a social media site, an activist group, a widely used software project, a lot of things. Because if you haven't, then you may not understand why we even use options 2-4 in that list above. But if you have stewarded responsibility for an institution, then you've had to make decisions about complicated situations, balancing among different goals.

  • Should we host our event at a clean, accessible space that's near transit and available for free, even if it's run by a company that does some stuff we're opposed to?
  • Should we nurture new leaders by giving them more responsibility and autonomy, even if they're making mistakes that might drive other participants away or cost us money?
  • Should we provide transparency by telling banned users why we're banning them, given that some of those users will use that information to make new accounts and sneak around our restrictions with subtler bad behavior next time?
  • We already signed an expensive contract with a venue to host an event during a certain weekend, and then found out it conflicts with a religious holiday, or a newly announced general strike. Should we cancel?
  • We fired, banned, or otherwise disciplined someone who is now publicly telling lies about why we did that. Should we respond, in private or public? If so, how?

Home versus coalition

And this tension gets harder if you want to do any coalition-building! A couple of years ago, probably via Camille E. Acey, I found and read Dr. Bernice Johnson Reagon's "Coalition Politics: Turning the Century" speech from 1981. As her site summarizes. "she drew an important distinction between the safe, home-like space that those challenging the status quo may need to bolster themselves and to help define their work, and the challenging, stretching, and often uncomfortable space of coalition-building." (I wish I'd read this and incorporated it into my phrasing before I gave speeches on the concept of "inessential weirdnesses" in open source outreach efforts.)

Dr. Reagon writes that you have to distinguish between comfortable, womb-like home spaces, which feed your needs, and uncomfortable coalitions, which you must feed. We ought to maintain both, but we need to recognize which kind of spaces we're in or which kinds of spaces we're building. And we need to recognize when what used to be a refuge has turned into a productive but often uncomfortable coalition, because that means we ought to shift how we operate, and make sure to build new refuges elsewhere to rest.

And we so often don't make that distinction. Which means we have conversations like this:

Person A: We are trying to make a safe space for [group x], which means not having to put up with [behavior z].

Person B: But we need to partner with and educate [group y] which means putting up with some amount of [behavior z] while we're in the process of teaching them not to do that.

Which is why it is good and necessary to have different groups, different spaces, with different shapes and goals. Some are informal and small and nameless. And they can sometimes do the best womb-like nurturing, while larger, more structured groups can be better at advocacy. To quote Marge Piercy's poem "The Low Road",

...Two people can keep each other
sane, can give support, conviction,
love, massage, hope, sex.
Three people are a delegation,
a committee, a wedge. With four
you can play bridge and start
an organization. With six
you can rent a whole house,
eat pie for dinner with no
seconds, and hold a fund raising party.
A dozen make a demonstration.
A hundred fill a hall.
A thousand have solidarity and your own newsletter;
ten thousand, power and your own paper;
a hundred thousand, your own media;
ten million, your own country....

Theory of change

So often, when I find myself a bit mired in disagreement or misunderstanding with someone else over an activist issue, I ask: what is my theory of change, and what is theirs?

"Theory of change" means: You have three things.

  1. An assessment of the way the world is
  2. A vision of how you want the world to look
  3. A hypothesis about some change you could make, work you could do, to move us closer from 1 to 2
    1. and, preferably, a way to check that hypothesis every once in a while, to see whether you're making progress or should switch to a new one

As the Beautiful Trouble handbook points out, you might come up with a few different hypotheses! And that makes sense to me. Different people and groups do different things that work toward the same goal in different ways. A refuge and a coalition, for instance.

Asking "what is your theory of change?" can be as irritatingly condescending as asking "what underlying problem are you trying to solve?" or "what's your business model?" with the same underlying subtext of "I'm asking because what you are saying sounds nonsensical." So I generally don't use those specific words.

But I do find it's a useful thing to work out. Where does this person disagree with me? Do we have different visions for a better world? Different assessments of the world we're in now? Different beliefs about what needs doing?

And: are they seeking home or coalition? What do they think this place is, when they defend it? And if maybe that means I ought to go make another thing that serves my needs, at least I know that's where I should start next.

Read the whole story
sarcozona
54 minutes ago
reply
Epiphyte City
Share this story
Delete

Metaphor and Growth | Cogito, Ergo Sumana

1 Share

I'm partway through a wilderness medicine textbook from NOLS (formerly National Outdoor Leadership School). Something from the section on burns keeps ringing in my head because it's interesting in its literal truth but also figuratively resonant.

Burns that injure only the epidermis, or that go deeper into the dermis, are generally painful. But deeper ("full-thickness") burns penetrate deeper, injuring the subcutaneous tissue as well. And, although the surrounding area might hurt if it gets a more superficial burn, the deep burn itself isn't immediately painful, because the burn has destroyed the blood vessels and the nerve endings that would convey the news of the damage.

And so that means that there is damage we can take that is so deep that it cuts us off from immediately grasping how much we've been hurt, because part of what it does is remove our capacity to know that. And the recovery will be painful, and more than that, it'll be disorientingly painful, as the fresh shock of the pain will be a sign of the nerves growing back.

I'm middle-aged now so I have decades of my life to look back on. And sometimes I'm telling a story to a new friend, some story I've told dozens of times, and for the first time it occurs to me, "hey, that could have gone differently if that parent/teacher/etc. had taken better care of me," and I feel the fresh pain of new grief over tissue damage I incurred eons ago. The nerves keep growing, and that's a good thing, even if the cost of growth knocks me sideways once in a while.

The weekly MetaFilter free-conversation thread this week spurred me to finally write this down after saying it to friends in several conversations these past few weeks.

I love learning more about medicine partly because it's useful -- I like being capable of helping myself and others with medical stuff, especially first aid -- and because the human body is just a fascinating set of systems. And because it gives me these lenses, these metaphors, that I can try to apply elsewhere.

But also, it grounds me. There is this physical experience that is pretty much the same for everyone who has ever lived. My ancestors, thousands of years ago, got burns and had to treat them. They bled from cuts, and got headaches and blisters and fractures. We know stuff they didn't, but they tried. And throughout history, there have been people trying to heal each other, learning stuff, and I get to inherit that.

(Which makes me feel a little more patient about the delay in NYC releasing the public defibrillator data. Not published yet. I know someone who's working on it, though.)

Read the whole story
sarcozona
59 minutes ago
reply
Epiphyte City
Share this story
Delete

Light Pollution Map | Bortle Scale

1 Comment

We're in the process of integrating real-time aurora prediction tools (such as KP index) and observation layers into our map. This will help users in high-latitude regions like Northern Europe, Canada, and Alaska track and pursue aurora sightings.

Read the whole story
sarcozona
1 hour ago
reply
Light pollution is a tragedy and a failure
Epiphyte City
Share this story
Delete

ABCD Study omits gender-identity data from latest release | The Transmitter: Neuroscience News and Perspectives

1 Share

Data on gender identity are missing from this year’s data release of the ongoing Adolescent Brain Cognitive Development (ABCD) Study, published at the end of June, The Transmitter has learned.

The U.S. National Institutes of Health removed these variables—which the study researchers have collected since each participant’s first-year follow-up—“to align with agency priorities,” a spokesperson from the National Institute on Drug Abuse, which co-founded and co-funds the study, told The Transmitter.

“Sex and gender are two different aspects of an individual, and ignoring one of those completely essentially adds blinders to us as scientists,” says Amy Kuceyeski, professor of mathematics in radiology at Weill Cornell Medicine. “I think the decision to not release gender information that was so carefully collected is unnecessarily restricting exploration of how neuroimaging correlates map to sex and gender together and separately.”

The ABCD Study launched in 2015 to track the brain and behavioral development of nearly 12,000 children, starting at ages 9 and 10, over the course of 10 years. Data are collected annually at 21 research sites across the United States and include MRI scans, neurocognitive assessments, physical health examinations and surveys with the participants and their parents.

A working group within the study that focused on gender identity and sexual health “has been sunset,” but sexual health data will still be available, according to the spokesperson from the National Institute on Drug Abuse. The spokesperson did not answer questions about what will happen to the data that have been collected since the previous release. Each release is cumulative and includes data from the previous years; researchers can continue to access archived releases until their data-use certification expires. It is unclear if gender data will be removed from those prior releases.

The removal from this year’s release follows President Trump’s January executive order that denies the legitimacy of transgender and gender non-conforming identities and required U.S. federal agencies to “remove all statements, policies, regulations, forms, communications, or other internal and external messages that promote or otherwise inculcate gender ideology.” The National Institutes of Health began reviewing other human data repositories for gender identity terms in March, The Transmitter previously reported.

This is “an arbitrary decision that’s not based on scientific evidence,” says Nicola Grissom, associate professor of psychology at the University of Minnesota. “Pieces of data that have been collected and that represent the important reality of children’s development have been removed, erased from the record” because they do not “happen to fit with a worldview that is not consistent with reality.”

T

he ABCD data on gender identity were collected via surveys. The study participants answered questions about how much they felt like a girl or a boy, how much they wished to be a girl and how much they acted like a girl while playing; parents answered questions about their child’s sex-typed behavior while playing and if the child spoke about having gender dysphoria, or feelings of distress that their gender identity does not match the sex they were assigned at birth.

It’s important to collect gender identity and expression data because these factors are part of the reality of adolescent development and can influence young people’s mental health, Grissom says. Excluding those variables counteracts “the goals of the study—which is to try to understand adolescent brain development—by pretending that some aspects of adolescent brain development don’t exist.”

Leaders of the ABCD Study expressed the same sentiment a few years ago. It is “essential” to study “the complexity of gender in adolescent development,” members of the study’s gender identity and sexual health working group wrote in a 2022 paper. “Gender is important across adolescence and advancing understanding of individual differences in gender development will broaden our understanding of adolescent development.”

Elvisha Dhamala, assistant professor of psychiatry at the Feinstein Institutes for Medical Research, previously used the ABCD dataset to study sex and gender differences in the functional connectivity of the brain. Now that the gender data have been removed, her group is focusing on sex differences instead. Many of the reasons why it’s important to study gender differences in brain development are likely still unknown, Dhamala says, because they haven’t been studied much in the past. “Once those data are gone, we just won’t be able to look at [that question].”

Read the whole story
sarcozona
6 hours ago
reply
Epiphyte City
Share this story
Delete
Next Page of Stories