The Both/And of AI and Trans Health w/Rebecca Sanaeikia
Episode description
In this episode, Élaina talks about the ethical challenges of using AI tools in healthcare provision for transgender people with philosopher and bioethicist Rebecca Sanaeikia. They discuss the different versions of “top down” versus “bottom up” ethical strategies and the tension between needing more data on how trans people access healthcare and wanting to keep trans people safe.
How to reach Rebecca
https://beccasanaeikia.weebly.com/
Twitter: https://mobile.twitter.com/logavaguy
Linkedin: https://www.linkedin.com/in/beccasanaeikia/
And you can email her here: rebecca.sanaeikia@gmail.com
Texts mentioned in the episode (All links are affiliated to Bookshop.org UK and any purchases made through them will generate a small commission that helps to support the podcast):
This Arab is Queer: An Anthology by LGBTQ+ Arab Authors, ed. Elias Jahshan
Subscribe to Philosophy Casting Call and leave it a 5-star review wherever you can!
Follow Philosophy Casting Call on Twitter and Instagram @philoCCpod
Read the full episode transcripts at www.elainagauthiermamaril.com
Support the podcast by becoming a monthly donor on Ko-Fi.com
Follow Élaina on Twitter @ElainaGMamaril
Transcript
Élaina Gauthier-Mamaril 0:15
Hello and welcome to Philosophy Casting Call, the podcast that features underrepresented philosophical talent. I’m Élaina Gauthier-Mamaril, you host and producer. Today, I have the distinct pleasure of sharing with you my conversation with Rebecca Saneikia. Rebecca is a philosopher who works on the ethics of AI usage in trans healthcare in the US, or healthcare for gender diverse people. We talk about her journey into philosophy from totally different fields, how she links her identity as a trans woman and her practice of interdisciplinarity, as well as the double-edged sword that is algorithm usage. This episode fits in really well with our previous discussions of ethical data collection and trans healthcare, so maybe check out my interviews with Hannah and Matthew once you finish this one! And now, on with the show.
Hi, Becca, welcome!
Rebecca Sanaeikia 1:13
Hi, thank you for having me.
Élaina Gauthier-Mamaril 1:15
Thank you so much for being here. Would you like to introduce yourself to the listeners?
Rebecca Sanaeikia 1:21
Sure. I'm Rebecca Sanaeikia. I'm finishing my second year PhD studies in at University of Rochester in Philosophy Department. I'm focusing on Trans Justice in healthcare, and I'm interested in bioethics, public health ethics, feminist epistemology, especially trans philosophy.
Élaina Gauthier-Mamaril 1:46
And all of that sounds very much like applied philosophy. You mentioned feminist epistemologies. And as someone who also has a philosophy background and works in bioethics and things, it's not always a given that people do philosophy before doing bioethics. So how would you describe how interdisciplinarity works in your practice?
Rebecca Sanaeikia 2:13
You mentioned this before; this is a very good and complicated question. I guess, even before studying philosophy, I studied architecture for my bachelor's degree. And that also is like an interesting interdisciplinary kind of major. The way that I see an interdisciplinary work in general is like, being trans, not finding a box, and always try to ways for yourself in any way that you will, in any way that you go, Oh, yeah! And for your audience, I am a transgender woman, if I didn't introduce that. So yeah, I think the way I see it is that I was interested, I got interested in philosophy after getting my bachelor's degree in architecture, working for some firms, and I realised that I have so many questions, and I need answers. And of course, I didn't find any answers! I found more questions in philosophy. But what kinds of questions is varied, it started with philosophy of art because I was an architecture major. So that was a way in the door for me to get into philosophy. When I got interested in metaphysics and more abstract questions, I studied philosophy in Iran, I'm from Iran.
Élaina Gauthier-Mamaril 3:40
In Tehran?
Rebecca Sanaeikia 3:42
Yes, I was born into Iran. I studied philosophy at the University of Tehran. So yeah, I came here in Rochester, New York, two years ago to continue my PhD. When I came here, to the USA, I realised that there are so many real life issues that I need to focus on instead of just doing an abstract and theoretical philosophy. There are so many important topics that I couldn't get my mind out of it. Like, when I go, especially for me as a transgender person, when I go for getting gender affirming care or any other type of medical care I face so many discrimination sometimes. And also, there's so much lack of research on transgender health in general. And for me, it was it became a real big question that how we should think about alleviating probably a discrimination against transgender people. That was just the door for me to find about bioethics find about feminist by a feminist epistemology and stuff like that.
Élaina Gauthier-Mamaril 5:03
It's interesting that you started in architecture, which is, as you said, interdisciplinary in a way like it is related to art, but it is also related to design. And, you know, a large part of architecture is very functional things, as well. So, you seem to always have been in that in between world of building in the world, but also theorising? And do you find that your perspective on art and art practice informs your current approaches to doing philosophy?
Rebecca Sanaeikia 5:36
The thing that affected me the most was the architecture that I did itself. Before that, it's interesting for you to maybe know that I studied with studying civil engineering for one year, and then I changed to architecture, which was more interdisciplinary and more interesting for me, I did computational architecture, and I was interested in using algorithms in architecture. And that's why when I'm working today, I think about the AI and the usage of AI in healthcare, and how we can alleviate discrimination by using that.
Élaina Gauthier-Mamaril 6:16
Yeah, I saw in the work that you sent me that you are working on this idea of how algorithms can commit certain kinds of injustices against transgender people and what you call gender diverse people as well. One thing that I found particularly interesting was this idea of when we approach transgender care or health care, it's often all lumped into one but there are distinct things that happen when you are gender non binary, and you're having to put down that you are in a binary or present that you are in a binary in order to receive certain gender affirming care. And so you know, when we talk about transgender health care, it's not just one size fits all for everyone.
Rebecca Sanaeikia 7:05
Yeah, that's exactly what I was thinking about whenever I talked about being trans. So the way that maybe human minds work is that we always try to use some categories to identify people. And that's actually good for us to do so in some situations. But the problem is that whenever it comes to gender identity, it's very diverse. It's where it has so many varieties and it's hard for someone to just, it's wrong and morally wrong for someone to categorise a person in a box when, for example, someone identifies as trans feminine, then it is wrong to assume certain characteristic of that person, because gender identity is another thing and gender presentation is another thing and how they just present themselves, how they navigate gender presentation is something else. I had an example of a trans feminine person that wanted gender affirming care, like hormone replacement therapy, but wanted also top surgery means they didn't want breasts. But that's not typical in healthcare. So when it comes to like, trans feminine people, they always think that, okay, this is the norm, this is the body that is transplanting and wants. And if you don't want this, so probably you're not transforming them. And that's when trans people or non binary people in healthcare face some dilemmas that how we should identify themselves in order to get the care that we need, which is discriminatory.
Élaina Gauthier-Mamaril 8:50
I can find resonance with some of what you're saying, from a disability point of view, when sometimes you there's a lot of discussion within the disability communities about having to perform a certain type, to perform your disability to various degrees in order to get and receive health care. You know, sometimes you have to be the good patient and play down everything. And sometimes you have to really exaggerate though the pain: "Oh, yes, it affects my life. Oh, yes." Like, even if that's not your current experience that specific day, but because, you know, depending on the care provider, this is the only way to access it. But can you speak more on what you mean by when you say it's morally wrong? Because some people make a distinction between ethics and morality. And from our previous discussions, it seems like you're really leaning into a kind of normative moral stance.
Rebecca Sanaeikia 9:46
Exactly. So first of all, I should say that when we talk about discrimination, not all of the discrimination that we use or wrong, in the sense that I will talk about. So discrimination is just comparing two things and just try to maybe give some advanced things to other, for example, in school teacher grades students, and based on that decide something, someone wants to hire people. And this is a discriminatory act, it is discrimination, they are trying to choose one person based on some merits. So the puzzle arises when people think about when it comes to play that we should say it is wrong. We should not do that type of discrimination, so yeah, I, when I say it is morally wrong I mean, it's not permissible I mean that there are some situations that is not permissible. And this is different than other types of wrongness. So for example, there might be some acts of discrimination that are harmful. And for that reason, they are wrong, but it's not because they are this nation that is wrong. Although I should mention that some discrimination theories talk about harmful, harmful theories of discrimination. But the way that I see it, which is borrowed from Deborah Hellman's view, is more about the normative way that we see wrongness.
Élaina Gauthier-Mamaril 11:27
Do you have any examples?
Rebecca Sanaeikia 11:29
So she talks about demeaning ways of... So she says, discrimination is wrong when it is demeaning. And by demeaning, she means that then we have the position of power, and we are doing that acts towards a socially salient group. In my case, for example, transgender people who have been marginalised, who have been suppressed in historically. So, for example, if someone is hiring someone, and they do not hire a woman, this act is discriminatory. And it's through a case of wrongful discrimination because they are doing that because they are a woman because of sexism. But then again, it's gets tricky, because sometimes we should think about intentions also involved in this kind of scenario. So, for example, there is a man and there is a woman and that man is merit-based, but like the married days are better than, for example, the woman that we are hiring, so our intentions are good, but we are not hiring that moment. And the dilemma arises that is it a wrongful discrimination that we are doing or not? And it's a really hard topic to think about.
Élaina Gauthier-Mamaril 12:51
Yeah, I mean, I have discussed that with my students when I teach feminist philosophy a lot. And the whole idea of merit is also very complicated. Because for, as you say, like historically, systems have been created to uphold a certain type of person who is worthy to have a job or is worthy to have a position of leadership or all these things. And so sometimes it's not so clear cut, like, is the merit really appropriate to the current situation? Or is this an entrenched way of viewing our ideal candidate? And so sometimes, you know, you might discriminate against people because they don't fit this particular version of what you think a good employee would look like, or a good doctor would look like etc.
Yeah, I agree completely. And one thing that I like about Hellman's account, which is well, I should mention that when we talk about when I am talking about discrimination, I am mostly talking about the discrimination laws, when it comes to to a situation that we think something is discriminatory and not permissible to do so. And it's a wrong act. What I liked about Hellman's account is that she uses exactly the phrase of "punching down", which we hear a lot, and I kind of resonate with it. So yeah, when you are punching someone down, it is wrong to discriminate like that.
I want to link this a little bit to your work on algorithms. And when you decide about what is the purpose of, for example, collecting data. So, we started this conversation and you said there was a lack of research in transgender health care. And then someone might be like, "Okay, well, let's go and let's collect data for insurance claims and all of these things". But you have an argument as to why that is, if not wrong, then maybe problematic. Could you explain?
Rebecca Sanaeikia 15:02
Exactly. So yeah, the way that I've worked with some of my other colleagues about this topic is that it is justified to do so. But we have to be careful that there are some cases that if you do use algorithms to identify transgender people, it's wrong. So mainly, I should maybe say what people do, usually, people design algorithms. And to identify transgender people, for example, in insurance claims, and they are having some criteria like using, if they are using gender affirming care if they went through gender affirming surgery, or if they are getting a surgeon, for example. And we might be, we have to be careful that by putting transgender people in the box as of just these characteristics, is just reproduce the boxes that I was talking about that you don't want to do. So although it's justified, it's a short term solution, because we lack so much knowledge and it's good for us to just, it's needed for us to do so because there are so many problems with collecting database on transgender people, because some of them cannot do not want to identify themselves as transgender because of safety reasons. And the other is that if they identify themselves as transgender, they are not the people that we are looking for in our researchers, we are looking for specific people who are getting gender affirming care. And we want to see how these care affects the other time the other categories of health. For example, what is the difference between trans women's breast cancer and ces women's breast cancer? And we should be careful because not all trans women are using gender affirming care, or not getting oestrogen, oestrogen. So yeah, but we should be careful. It's good. It's very helpful, even for me. And that's one of the reasons that I'm doing this research because I need to know I need the data, I need to know what's happening to my body when I'm using a surgeon. But at the same time, we have to we have to be careful about the consequences.
Yeah, you raise interesting points about how we could look at it from a certain population seeks access to health care. But that's way too simplistic, because there are so many factors that play into that. And whether or not people feel safe to identify as transgender or in a medical setting, whether or not having something on their file sending, saying that they access gender affirming care will affect other unrelated health care that they receive in the future who shares those databases. And yeah, I think your emphasis on looking at US law is is interesting in that case, because it is a very unique landscape of how all of these things, it kind of really shows how anyone, when they make a decision to seek medical care is going in with a lot of identities and a lot of issues. And this really shows, you know, some people like trans people will have to decide whether like I have gender affirming care, or I deny myself access to cancer treatment in the future, because someone will discriminate against me.
Exactly. You're talking about some of the criteria of like, the Belmont Report, which is like just informed consent. If I really want to have an informed consent, if I really want to give my consent to go to some gender affirming care, I need some information. But there is so much lack of information about transgender health in general.
Élaina Gauthier-Mamaril 19:15
So it sounds like it's like navigating a liminal space of you know, it's a lot of opportunity, but also a lot of danger in how we tread and how we advance.
Rebecca Sanaeikia 19:27
If someone uses these algorithms, to just identify trans people and then just not giving access them in some scenarios, and it's so much harm, don't doing to our transgender people, so we might be careful about that.
Élaina Gauthier-Mamaril 19:44
So can you tell me a bit more about how you're using feminist epistemologies as a methodology in your work?
Rebecca Sanaeikia 19:49
I particularly think about epistemic injustice. There's two ways of epistemic injustice and epistemic justice is just... When we talk about epistemic injustice we are talking about giving less credibility to some socially salient groups. In my own arguments, I am thinking about transgender people and by transgender people, I mean whoever that identifies as a non cis person, so non binary people, transgender women and transgender men, trans masculine trans feminine. Then, there was a good example of trans, these types of epistemic harm, which was coined "trans broken arm syndrome". This refers to whenever a transgender person goes into clinic, the physician identifies a problem as the result of their hormone replacement therapy, and they don't investigate further. And this is a similar pattern when it comes to cisgender women in health to when they go to clinical practices when they face physicians, if the physician for example is a male, then they probably think that their pain is not as real as men or is not as important because historically just the male body was the centre of attention was the centre of research. So yeah, I think about these and these are, these have some impacts, like, when it comes to a self identification, when it comes to gender identification informs it will be like testimonial, it will be some part of testimonial injustice, which means, which I call which is the dots and call it testimonial smothering which means I have the concept, I know what to use to identify myself, but because the person who is in front of me the audience, wouldn't give me the credibility or do not believe me, then I will don't I don't use that concept. And instead of that, I identify myself as somebody that something else. For example, when it comes to firearms, as we talked about before, I will identify myself as a transgender person, even though I don't identify myself as transgender person in general, but identify myself as transgender because I know that they want me to identify in order to get them requirement care, or in order to inform them that I went through gender affirming care.
Élaina Gauthier-Mamaril 22:28
Yeah, that's really interesting. And, again, it's just so many things interplay, like anecdotally in my life, like I have trans friends who report you know, bettering or worsening of their chronic conditions when they started hormone replacement therapy. And that is not something that is investigated because again, as you say, like the male body is the default and anything that is deviant from that. We have, like a lot of issues have getting people to pay attention to like, female cis bodies, and then now you're coming in saying, well, there's actually infinity variations on this.
Rebecca Sanaeikia 23:09
Yeah, exactly! So I'm talking about not only male bodies are centred in healthcare, cis bodies also have been centred in healthcare. And there have been so many structural injustice is towards trans people or trans bodies. So we have to start researching. One way is just by using AI. The other is just by keeping maybe thinking about variation of how we should collect like gender identity, or how we should think about if the current sex assigned at birth, like the two sexes, that exists is enough, because we are excluding, for example, intersex people. So there is so much to work on.
Élaina Gauthier-Mamaril 23:58
So, I come from a perspective where ethical AI makes me very suspicious, this idea that you could... Because there's always a human being behind creating the algorithm. And then once it's enacted, like there are lots of evidence of AI, medical AI, not picking up darker skin tones, for example, when identifying certain things. So what do you think is the future role of AI and transgender healthcare?
Rebecca Sanaeikia 24:31
That's the big question. I wish I knew the simple answer for that question. But I think...
Élaina Gauthier-Mamaril 24:38
Maybe ideally, like what would you like to see?
Rebecca Sanaeikia 24:41
I think it works hand in hand when it comes to transgender population, because systemic injustice exists in society, no matter if we think about AI or not. So when we think about AI, I think we, our approach should be top down. And there's like, there are two approaches. So either we can think about ethical things and then implement them in AI or either or we can instead use AI to think about algorithms and then go to the top and make AI ethical. And I think when it comes to population, like transgender people, we should focus our top down. And I'm emphasising on that, because I think the first should work so much on our society on how the systemic discrimination against trans people are currently worked, and built into healthcare instead itself. So yeah, it's very hard to just ignore that and think about that AI will magically solve all the problems.
Élaina Gauthier-Mamaril 25:57
So just to clarify, when you say top down, do you mean, we need to reevaluate or revolutionise the framework in which we approachhealthcare in general?
Rebecca Sanaeikia 26:11
Okay, exactly. So that's what I'm working on my dissertation to. So I'm thinking about the ways in which we are thinking about gender in medicine in some contexts, we don't need to use gender identity in others, we have to use gender identity. One example is the AI that we talked about how we should identify some people because we lack knowledge about transgender health. But in some scenarios, the gender identity itself, it doesn't do much for us, and it's even do harm for transgender people, or intersex people are other people that are being excluded from this research. I think the way that I talk about top down is that well, we should think about from the structural and institutional level, mainly institutional, institutional level, when we are talking about transgender health. Think about the ways in which we can talk about gender in healthcare system, and then how we can implement those ethical considerations in AI.
Élaina Gauthier-Mamaril 27:18
And how much or how little do you see engagement with actual trans people in the development of this framework?
Rebecca Sanaeikia 27:27
That's a good question. I think there's so much research must be done, at least in other disciplines, like sociology that we know more about transgender population in general. But I think, well, there are so many movements, activity activism going on, even for Transgender Health. So the way I see it is that we should be visible, as visible as possible, and try this to be the presentation that is lacks now in the healthcare system, for example. And, of course, the knowledge that we have to live, the experience that we have, would be so much helpful for each level of these things that I was talking about. So I am working on these issues, but one of the reasons that I can work, maybe I can contribute better... It's not good to say better, but I can contribute in a way that maybe a cisgender person cannot contribute is that because I have the least lived experience of transgender person, I have the knowledge, I know the trans culture, for example, some people might not think of transgender, there are like trans dominance, that's like they want top surgery. So I think the solidarity actually in our community and come to talking about these things are very important. And the other probably population can give us some places to talk about these issues. Because there are there are so many transgender people rising and talking about transgender health, or other problems in general, about trans individuals.
Élaina Gauthier-Mamaril 29:21
Yeah, I know, it's very interesting because I ask, because it my understanding of feminist methodologies is much more I bottom up or like a grassroots version. But as a philosopher, I'm also very sympathetic to the need to assess the structure. And what I'm trying to do in my work right now is trying to... How do you manage a dialogue but so it's not one or the other? It's both/and.
Rebecca Sanaeikia 29:49
Yes!
Élaina Gauthier-Mamaril 29:50
And this idea in the disabled community of "Nothing about us without us" and how much harm can be made from like well meaning people who are trying to create services that have a charity model of disability, for example, that is very dehumanising. And so that's why my questions were kind of like, how do you think about engaging people? So it's not just like fancy philosophers in a tower I thinking about how do we redo the healthcare system?
Rebecca Sanaeikia 30:17
Yeah. Yes, that's not the correct way that I was thinking about. The way that I was thinking about AI was exactly, actually, that's helpful, I should think about it more... About the top down method, I just was thinking about the exact wrongness against trans people. We should use these criteria when we are trying to design our algorithms instead of just like, using algorithms that is trying to learn for itself, because definitely will make some stereotypes about transgender people. It's not aware of the transgender population. Maybe in some scenarios it can be useful, but I have my doubts, too. But yeah, but in order for us to have that top down approach, definitely, you're right, I think we have so much work to do. So, actually, also in my research, I have separated two distinct things: gender in an institutional level and gender in an interpersonal level. In an interpersonal interpersonal level, I think, exactly what is important is gender. A physician should and must think about the gender identity of a person, if they really need to share that information with the physician. If we are going, for example, to a mental health doctor, then knowing that we are transgender is important information for them to know. So yeah, I also think that it goes both hand in hand and I agree with you.
Élaina Gauthier-Mamaril 30:18
Is there anything I haven't asked you that you want to talk about?
Rebecca Sanaeikia 31:48
No!
Élaina Gauthier-Mamaril 31:52
Okay! Well, I'll ask you, then, what are you reading right now that's giving you joy?
Rebecca Sanaeikia 32:07
Yeah, I actually prepared for this. The audience cannot see it, but I read i"This Arab is Queer". Yeah, it's an anthology by LGBTQ+ Arab writers. It has such some short essays about their experience in their own communities and their own countries. What I like about this book is that I find myself in them because we have some similar culture. I'm from Iran. The difference is that Iranian people are not considered Arab, so we don't have representation in that book. But at the same time, the culture is the same. So I'm loving and enjoying this reading each part of this book.
Élaina Gauthier-Mamaril 32:54
That's great. And is there anywhere that people can find you and or your work on the internet?
Unknown Speaker 33:01
Yeah, I have a website beccasanaeikia.weebly.com They can also find me on Twitter @logavaguy; if you search me I will show up. I am also on LinkedIn and they can all find out this information on my website too. And they can also email me I will love to just talk about these issues that I talked with you recall at beccasanaeikia@rochester.edu
Élaina Gauthier-Mamaril 33:29
Right. And I'll put all of that in the show notes so people can find that easily. So thank you once more for coming and talking to me. This was really, really fun. And I hope you have a good day!
Rebecca Sanaeikia 33:41
You too. Thank you very much for inviting me!
Élaina Gauthier-Mamaril 33:53
Thank you so much Rebecca for a fascinating conversation and I look forward to following your work as it develops. As usual, you can find ways to contact Rebecca in the show notes. If you want to support Philosophy Casting Call, the best way is to rate the podcast and to leave a review on Apple Podcasts or Spotify. Also tell me if you use Philosophy Casting Call in your classrooms, cause I would love to know! If you are in a position to donate, you can also become a monthly supporter on Ko-Fi.com You can follow the podcast @philoccpod on Twitter and Instagram and all the transcripts live on my website at www.elainagauthiermamaril.com. Until next time, bye!
Transcribed by https://otter.ai