Hi and thanks for subscribing to my newsletter! The breakdown is as follows: a personal essay on top of the letter and some more concrete life updates, current media favorites and a recipe at its bottom so feel free to skip to whatever interests you. (Please feel free to hit the Reply button at any time, for any purpose.)
I know these letters have been somewhat off-schedule lately, but my evenings have been heavy on work and my weekends oddly filled with social engagements for the last few weeks which pretty much eradicated all of the time I would have usually blocked off to write. This letter is also not topical to anything currently in the news since I sketched it out a few weeks ago, before the most recent wave of violence and controversy in the United States. Honestly, I am somewhat relieved to have had this text to work on as more and more terrible news have been coming in since it gets hard to process so much fear and hurt day after day. I hope you are holding up as well, especially if you live in the US – I care for you much more than I could be scared or saddened by something a politician says.
VIRTUAL IMAGE*
One of the most famous hip hop groups in Croatia, called TBF, counts a song about commander Data from the classic television series Star Trek: The Next Generation among their most recognizable hits. They’re a goofy band with songs that often poke fun at both pop culture and local politics. Data is an upbeat song that leans heavily into the nerd status associated with its subject, shouting-out other characters from the show and referencing electric sheep for an extra touch of convincing sci-fi literacy. There is a verse in the song that translates to something along the lines of ‘and he always makes me think of the same question, what do we have to do to be human?’ .While this quite accurately reflect so many of Data’s storylines on the show, he is after all perpetually chasing some ideal of Pinocchio-like realness, it also summarizes the main role of robots as metaphors in media fairly succinctly – by positioning the robot as something not quite human and then allowing them to go on a quest for humanity and human-ness, it is implicitly defined what it is that is human, normal and regular.
The role of a robot, such as Data, has often been interpreted as a stand-in for marginalized or in some way other-ed individuals in our society. Similar arguments can be made about aliens, witches (here there is some historical background to further prop up that interpretation) and, most famously, mutants (even for complete X-Men novices, pieces such as the 2017 Logan movie that grapples with immigration in a pretty un-subtle way are powerful examples). In other words, instead of tackling issues of race, gender and gender identity, sexual orientation etc. head on, creators often opt to populate their worlds with characters that are different in visually striking ways that then power a metaphor for some contemporary type of marginalization and oppression. As this approach provides more opportunity for the creator to formulate the rules and context of such ill-treatment, this manner of discussing structural tensions within the real world society can be rather risky, especially once one acknowledges not only how different experiences of different marginalized groups can be but also that real people can and do carry various overlapping identities within them. It is quite a challenge to paint a truly intersectional picture of an experience when the only qualifier that sets a character apart from the rest of his crew is that they are, for instance, blue.
Data’s character however, as a function of his longevity and popularity, is a fairly complex and rich stand-in for a person that has in some way been deemed different and categorically other. Beyond disputes of his humanity and his legal status i.e. his autonomy over his own body (something real life marginalized people are often very familiar with especially as healthcare becomes more and more inaccessible and certain types of care and medical interventions heavily legislated), Data’s story includes brushes with romance and sadness, a never-ending quest to tell a good joke and an evil twin that has been corrupted by too much emotion. In this way, Data is something of a poster child for the kind of respectability politics people of color, women and gender-non-conforming folks** are often subjected to: his failings to adhere to the social norms the rest of the crew is proficient in are acceptable and endearing as long as he does not get too loud and too emotional. Further, his need to learn how to be human and his remarkable expenditure of energy on trying to fit in highlight that assimilation is much more important than self-sufficiency or independence. Viewed in this light, it is almost sinister to notice that Data’s best friend on the show is Geordi La Forge who happens to be a person of color and differently abled. While one rarely encounters media where two robots just hang out as friends instead of palling around with humans, unless they are plotting a violent rebellion, here is an example of two othered, or at least capital-D different, characters spending time together and bonding. A very favorable reading would be that this is a metaphor for the kind of chosen family queer individuals often construct for themselves following a rejection from their biological relatives. A more down-to-earth take is that in a crew of normals, the two avatars for marginalization have to turn to each other because they don’t have any other means of seeing a reflection of their own experiences.
Star Trek: The Next Generation first aired in 1987 when the idea of a sentient robot, or at least a machine that can talk back was truly something futuristic. Today, there are five-year-old’s that have to be taught to say thank you to their parents’ Alexa. As the meme goes, we did not get flying cars and this is most certainly not the dystopia most of us thought we were signing up for, but our everyday routines are populated with more and more machines hiding behind phrases such as ‘personal assistant’, ‘the Internet of Things’ and ‘neural network’. Algorithm is a buzzword and even the most tech-clueless of commentators have learned to complain about it. I ordered a silicone donut mold off of Amazon recently and when the company lost it a few days before the dinner party I was planning to make (baked, vegan, pumpkin) donuts for, instead of calling an automated service I messaged with a representative on their website until I was promised a refund. Most certainly the representative was a robot following a fairly simple script meant to assess the best course of donut pan action. Scrolling through social media I see former classmates using machine learning for particle physics research; there is an ad about weeding out fake and malicious users from the website on top of my Facebook feed. When my hands are wet from doing dishes I still ask Siri to tell me how long until my dinner is done roasting.
In the same way that writing stories about robots has in the past served as means of defining what being human means, and what sort of transgressions land one in some other category, the way we design and treat the real robots among us today reflects our beliefs and biases about other people. For instance, most personal assistants have stereotypically female voices and female names while AI tasked with playing games or competing on game shows get gendered male. It is remarkable that we feel the need to assign gender to robots at all but the split in the distribution of such an artificial assignment is certainly in line with traditional gender roles many continue to claim we have transcended already. It is mildly terrifying to notice this, and I have personally become more and more weary of the kind of designer that gives gender to a robot – why are they so invested in propping up the power asymmetry system that the institution of gender is? Even the original Turing test involved a gender determination.
Further, as so many of the new technologies rely on learning from large datasets before they are able to make predictions or ask questions or take action, it is hard to imagine that some amount of bias is not hardcoded into their ‘personality’ shaped by those datasets. To an extent, such design could be improved upon by coding in some type of a consistency or a sanity check where an algorithm or an AI could look at what it has produced and then cross-reference it with some criteria for certain types of bias. However, humans would have to decide on those criteria thus leaving room for other biases to creep in. A biased algorithm is thus almost more human than the idealized machine that is completely impartial and that we often naively think of as being some sort of a magic solution to bad human decision making. However, while a human being can independently learn to be self-correcting, it is not clear whether an AI can do the same.
As robots continue to advance, the more tangible elements of design also become more of a concern. If a robot is strong and supposed to participate in combat, will it look like a large man? In such a case, relying on gender norms could prevent us from discovering the best design for the task, if we fail to see beyond the association between masculinity and strength. The issue of gender is still somewhat simple to visualize but with any humanoid design, a number of other physical traits are likely to come into play and typically no physical trait relating to a human body is value neutral. Given a chance to pick and choose these physical traits with some task in mind we may not always intuitively settle on the most efficient one. The issue gets even more complicated when considering the possibility of interfacing robots and humans. Will robotic limbs and similar enhancements be used to fully eradicate the notion of differently abled people? Will those that cannot afford robot parts be othered and marginalized even further? Will the process of becoming a cyborg for medical reasons be sensitive to race and sex? The history of testing drugs such as the birth control pill and the general practice of disregarding women in medical trial indicates that some bodies are seen as more important than others. It also suggests that some bodies are assumed to be expendable. It is therefore reasonable to fear that not only will the design of robots only rely on certain bodies as a blueprint but that if robotic fixes to medical issue become the norm many will be either unable to afford it or simply not compatible in the most basic Lego-like sense. (I have had LASIK and I have an IUD; I am invested in being bionic.)
As robots and AI become more and more prominent in our every day life and we spend more and more time grappling with the very ethical and very practical concerns that designing and using them brings, this relationship quickly starts to become bidirectional. On Star Trek, Data was one of the fan favorites and an almost aspirational character. In line with Gene Rodenberry’s highly utopian take on the future that the show brings to life, Data’s rationality is framed as empowering him to always do good and be kind even though the viewer is often reminded that he does not have feelings. Humans are hotheaded while Data (again, unlike his evil twin) is not so maybe the viewer should aim to be more like that as well. AI sales representatives are endlessly polite and repeat the script until all options are fully exhausted so maybe human telephone operators should be similarly relentless. Alexa and Siri will answer your questions even if you don’t say ‘maybe if you have time, could you please’ so maybe other women in your life will do work without those caveats too. This analogy is crude and generates quick pushback once the robot part of it is real and not a TV character. In real life, we think, we know where to draw the line between the artificial and the real and only the latter, the real, the authentic, the biological, is granted the privilege of a second thought and politeness.
Yet, the artificiality of a robot is one of the reasons why it can function as metaphor for certain types of marginalized people or give them someone and something to find themselves in in the media. The fact that a robot is constructed and can be re-constructed or reconfigured is one of its most appealing traits when one is convinced that they need to change something about who they fundamentally are in order to fit in, be accepted or simply survive. We bristle at comparisons between how people treat Siri and actually human personal assistants because we think we value human authenticity, but the truth is that we only value certain types of authenticity and everyone else is invited to create an artificial persona. In the age of social media this is more pronounced than ever. The fact that being a social media influencer is a career and dating has almost fully moved online for a very wide demographic underscore how important it is to be able to design avatars of ourselves, to attach ourselves to some by default artificial simulacrum of what we might actually be like. This is as constraining as it is empowering.
Being given the liberty to construct ourselves to be who we want to be can be extremely powerful and elating. On a very basic level, this is why we have specific haircuts, wear particular shoes or give into or refrain from makeup. Queer folks have always embraced this, from their embrace of fashion to extremes of drag culture. Welcoming non-conformism as a choice is a defense mechanism that repackages the authentic self that is otherwise unacceptable. I keep thinking about Janelle Monae coming out as queer and having no sexual preference based on gender (something I have always felt as well), after building a career on a robot persona, Cindi Mayweather, woven into her music. Following the release of Monae’s latest, Dirty Computer, a lot had been written about how her constructed identity blends in with her as a person and part of the appeal of this record definitely lies in the fact that it is more blunt, more in-your-face, less robot Cindi and much more queer, black and proud Janelle. Ironically, the two have never been that different as Cindi’s transgressive love has driven Monae’s past albums, as her status as an outsider, an other and a rebel has been a key part of her character. What has changed in ten years between Monae’s debut EP Metropolis and Dirty Computer is the social norms, not the musician herself. In this way, the robot persona has been useful to her. A forbidden love between a robot and a human is after all something sci-fi and pop-culture consumers are used to, and they may not consciously recognize it as a metaphor for something that may upset them in real life. There is an episode of Star Trek where a love-fever-like intoxication takes over the crew and Data is seduced by the security officer Tasha Yar. The robot is treated like a heterosexual man, down to making sly comments about his functionality in that realm. His friend Geordi, human yet black and differently abled, however, is never shown having sex in space; compared to individuals at an intersection of some identities even robots are granted a fuller version of humanity.
At some point in the last few years, my boyfriend and I committed some time to HBO’s Westworld, a show that offers a much grittier and gorier (and since it is an HBO production, also a much more violently sexualized) treatment of sentient robots and their interactions with humans. We made it through the first season and sort of forgot about it afterwards. It is a visually compelling show with some great performances from some great actors but the pacing of the first season struck me as odd and the themes as somewhat unoriginal. Heavy nods towards the theory of the bicameral mind towards the season’s end helped its case but overall, I found myself wondering whether I could even be invested in a story of robot rebellion and self-discovery anymore, whether I could shake the feeling that all of these narratives have become stale. Yet, the importance of stories that essentially frame what human-ness and humanity are is evergreen and put into context of a world where robots are everywhere, and we are constructing and designing ourselves daily, they seem so much more important than what a TV network can put together to rack up views. Being desensitized to what’s on the screen is a pitfall that can lead to being desensitized to the way things are changing beyond it. As humanity of so many groups of people is being devastatingly questioned by figures of power, we should be critical of fiction that comments on it and conscious of how it affects our takes on the technology around us. It is too late to panic about an upcoming robot uprising – they’re everywhere already – but we can at least hardcode some compassion and freedom into it.
Best,
Karmela
* In ray optics, an image produced by a mirror or a lens is referred to as virtual when it seems that the light rays making it up would have had to come from behind the lens or the mirror. For ordinary mirrors all images are virtual – if our eyes were to see the same thing without the aid of the mirror, the object we are seeing would have to be positioned behind it.
** Respectability politics refers to the notion that marginalized groups only deserve equal treatment as dominant ones if they satisfy certain arbitrary criteria of appropriateness. In this way only voices compliant with the mainstream are given the benefit of the doubt. A typical example is judging group based on the way they dress so for instance, advocacy for racial equality would be seen as not worth being taken seriously if coming from a person in hip-hop-style street fashion as opposed to someone wearing a suit.
ABOUT ME LATELY
LEARNING: I’ve seen a number of really good talks and seminars in the past few weeks, most of which impressed me by establishing creative yet concrete connections between theoretical and experimental work. This is something that I mention often because as much as I hope I am growing as a researcher it still remains really important and, frankly, really mind-blowing to me that certain quantities can be measured or inferred from some cold chunk of metal in a lab. For instance, one of the talks I saw discussed implications of topological properties of materials on the way they interact with light. Topological properties are somewhat difficult to explain in lay-terms since they are typically defined with respect to an abstract mathematical space dual to ‘real’ space in which we generally think of objects as residing. In a way, this is exactly the punchline – topology is a priori defined in a way that does not necessarily have to lend itself to experiments (except maybe in the celebrated case of the Quantum Hall Effect). Therefore, being able to see any topological effects in real measurements in real space seems like such a great feat to me, even when I have heard of other examples before. To a large extent, these talks and presentations are also a good inspiration for my own, theoretical, work. As a graduate student I do not have full, absolute control of how projects I work on shape up, but I have been lucky in that real so far and have chosen to always be very vocal about chasing some connection with contemporary experiments. Incidentally, my group has made some steps in the direction of a new experimental collaboration recently. Given that it involves some very new yet very accessible materials I have been cautiously excited about it. I have also had a chance to meet with one of the many visiting scholars that pass through my university and they were quite excited about one of the problems I am working on (hopefully also at least somewhat experimentally feasible) even though it was only tangential to their own. (Full disclosure: studying ultracold bubbles is pretty niche even for a theoretical physicist) Five years into my PhD and five years after college, I am still getting used to my seniors taking my ideas and efforts seriously so explicit displays of enthusiasm seem quite validating. On the other side of that spectrum, some of my research sparked by another possible experimental collaboration in this past week landed me knee-deep in very dry and mathematical literature about K-theories and C-algebras and this has been quite a salient reminder of how much sheer fundamentals there are to learn and how much the mathematics degree I do happen to have should really be viewed as just as springboard into some much deeper material. And I was already having nightmares about singularities of Jacobi theta functions.
In addition to my two-to-two-and-a-half research projects I am also still teaching most days of the week and still organizing events whenever my schedule allows. I got some very positive teaching reviews at the half-semester mark and those felt very validating and honestly just plain good. At my university, teaching assistants receive virtually no training, so I have always been cautious and self-conscious about teaching. Having thought about it so much in previous semesters seems to be paying off now and I am more confident in my ability to help students more effectively than in the past. I hope I can keep that good momentum until the end of the semester, especially since the course I am teaching just made a pretty serious shift in topics. On the organizing front, I’ve helped put together a fairly successful event for International students that will likely be repeated next month, I contributed to the graduate employee union having a transparent budget and clean books for the first time in its history (hunting down transactions and using financial software is a nightmare but at least now I understand why anyone would want to recruit a physicist to a financial committee) and hopefully a few more events catering to women and gender minorities in my department will materialize before the end of the semester (in part thanks to some awesome friends committing their time to these efforts too). All of this work is either a distraction from my research or a drain on my free time, but I continue to feel that it is absolutely necessary and that I would be making a mistake by not at least somewhat committing to it when I can. With all the horrible things we see in the news over and over again, even the smallest acts feel like they matter, and I want to lend a hand in propping up as many people that might feel left out or threatened within the local academic bubble.
And yoga? I think I’m close to understanding what it means to ‘press thigh flesh into the thigh bone’ and ‘rotate inner arm to the outer arm’. I still can’t do a handstand or a headstand or a split, but I still enjoy the practice very much.
LISTENING: In addition to dipping into the three TBF songs that I often listen to when I am homesick, I started thinking about this letter upon hearing LeVar Burton talk about Geordi Le Forge’s importance as a person of color in the cast of Stark Trek on this episode of Gimlet's The Nod. It is an interesting episode to someone who grew up abroad like I have an did not get to experience Roots or Reading Rainbow but the hosts’ exuberance regarding all of these shows is quite convincing. Once I started thinking about robots, there were many great podcast episodes to think back on, including those dealing with sexbots in surprisingly illuminating ways like on Stuff to Blow Your Mind here and Flash Forward here, but this episode of Unladylike was likely my favorite. I am somehow never fully caught up on Unladylike and that is a mistake on my part as hosts Cristen and Caroline consistently produce relevant and layered stories about a very broad swath of feminist issues. The breadth and the distinctly contemporary feeling, compared to older and more stuffy feminist commentators one sometimes encounters in media explicitly marketed to women, really make this show important and worthwhile. On a somewhat more big picture level, this interview about the importance of algorithms from Inquiring Minds and this really great discussion of effects of cataloguing practices in libraries, taking the issue all the way back to Soviet Russia, on Lost in the Stacks both offer a more institutional take on some of the effects of smart and smarter technologies we now live with. And if you want to be really terrified about digital security, Breach is back with a new season which opened by tackling voting machines but the previous one, dealing with a massive Yahoo hack, will definitely do the trick as well.
As far as music goes, I’ve mostly been mesmerized by some of the recommendations that came up in a follow-up to an episode of Supercontext that discussed the Neurosis’ record Through Silver In Blood. Having been a big fan of a lot of ‘classic’ post-metal for a couple of years, I was familiar with the record when I listened to the episode but when I revisited I was once again struck by how raw and personal it is and how much it plays up just the right kind of grating loudness that I seem to take a comfort in. Both Tribes of Neurot (this record is probably a good companion to Through Silver in Blood) and Goodspeed You! Black Emperor (these two records were recommended by one of the Supercontext hosts) are absolutely mesmerizing but in very different ways and Sumac is just the kind of grating and heavy band I would love anyway (and their latest record is great).
WATCHING: My boyfriend visited for one of the weekends in October and between catching up with friends and very Illinois type activities like visiting a corn maze, we made just enough TV time to finish Netflix’s Maniac. As much as I was taken with the aesthetics and the clearly intentional absurdity of the show in its first two thirds, I became more and more disappointed with it neared its end. There is not much that changes with respect to the approach or the course of the story and there are definitely some quite striking (and strikingly gruesome) visuals in the later parts of the season but the show simply seems to saturate its camp-and-over-the-top-ness limit too early. As a viewer, even when I liked recognizing visual references and caricatured tropes, I just got fatigued by the time some semblance of an emotional climax was supposed to deliver its punch in my direction. Had it ended sooner or in a less prolonged fashion or had some of the storylines been edited just a bit more this could have been a great show and one that I would, as a former cyberpunk fanatic, appreciate very much. This way, I will probably mostly remember the bulky computers, Sonoya Mizuno’s triangle wig and glasses, Jonah Hill being so much better than I remember from bad Appatow movies and that one guy that literally fell apart. Even the ending disappointed me by playing into the most generic of tendencies the show exhibited instead of even just reaching for some other predictable but more interesting twist.
Additionally, on the weekend before Halloween I made time to go see a bad old movie at a local art theater. I really enjoy taking myself to the movies and the 1988 version of The Blob was a pretty good choice for doing so without having to engage in any sort of emotional heaviness. This is not a good movie nor is it a very well-made movie. It is mostly a collection of ridiculous 80s horror movie tropes glued together by the means of an alien invasion, military experiments gone wrong and a cheerleader-meets-town-rebel love story. The film tries to shock and show gore but to a millennial movie goer like myself these attempts seem both awesome and hilarious and consequently fail to qualify as jump scares. Disregarding one very uncomfortable ‘date’ scene (a good reminder to not get too sentimental for not having lived in the 80s), this sort of nonsense is exactly what a relaxing Friday night in late October should consist of.
EATING: This past weekend, I attended what was supposed to be a picnic for a group within my department but somehow ended up being just me and a friend eating lunch in a park on a warm and windy Sunday morning. She brought a baked pumpkin mac’n’cheese and hot chocolate and I brought the salad I am sharing below. It is a pretty predictable mix of greens and roasted vegetables given this time of the year, but that predictability also makes it into a solid staple and I have very successfully brought variations of it to family affairs along the lines of Christmases and Thanksgivings before. The dressing I am suggesting here is inspired by one of my favorite cookbooks, Gena Hamshaw’s Power Plates, but the nice mix of textures resulting from well roasted squash, crispy chickpeas and raw kale is really what makes it shine.
For 2-3 servings of the salad as a main dish and roughly 1 cup of dressing (i.e. you will have extra), you will need:
For the salad
1 medium butternut squash, seeds taken out, peeled and chopped into bite-sized pieces
1/2 bunch kale (6-8 leaves), washed, dried and torn into small pieces, discarding the tough ‘ribs’
1 ½ cups cooked chickpeas (or one 15 oz can, drained and rinsed)
1 tablespoon olive oil (or more)
Juice of half a lemon
1-2 teaspoons chili powder
Salt
1/3 cup dried cranberries (naturally sweetened if possible)
For the dressing:
2 tablespoons olive oil or tahini
1 tablespoon maple syrup or brown sugar
1 tablespoon Dijon mustard
1 tablespoon soy sauce or tamari
1 tablespoon apple cider vinegar
1 tablespoon balsamic or red wine vinegar
1 ½ tablespoons nutritional yeast (or parmesan if you eat dairy, or simply omit)
2 – 4 tablespoons warm water, for thinning to desired consistency
Cut the squash: cut of the top and bottom (with respect to the long axis of the squash) of the squash so you can balance it more easily, then cut perpendicularly at the point where it starts to round out. Cut each of the pieces in half vertically. The rounder part should contain all the seeds – scoop them out with a spoon. Peel the four pieces then cut into bit sized pieces. In the meantime let the oven heat up to 400F and line a cookie sheet with parchment paper.
Arrange the squash pieces on the cookie sheet so that they are not too crowded (if they are too crowded they steam in the oven and take much longer to get crispy and caramelized rather than mushy), toss with a good pinch of salt and olive oil. Once well-coated roast for 40 minutes, rotating the cookie sheet and stirring halfway through. Once done, the squash pieces should be lightly browned, and some should be visibly caramelized.
Massage the kale: once you have torn the kale into pieces, put them into a large bowl and squeeze the lemon over them then add a pinch of salt. Mix well. Squeeze handfuls of kale with your hands, rubbing the lemon juice into it until it looks wilted and some liquid has collected at the bottom of the bowl. Discard most of the liquid and let the kale sit in the fridge while your squash is roasting – the acid and the salt will continue softening it.
Crips up the chickpeas: get a non-stick or cast iron pan very hot and dry the chickpeas as best as you can with kitchen or paper towels. Pour the chickpeas into the pan and slightly lower the heat. Dry roast for five minutes or so, stirring often, until some chickpeas start to look darker and slightly charred. Ideally you’d have a big enough pan to have them sitting in a single layer. Once the chickpeas start to visibly char, sprinkle them with chili powder and stir until well coated (no streaks should remain). Continue dry roasting for a few more minutes or until done to your liking.
Make the dressing: combine all of the ingredients in a jar, put the lid on and shake until well combined
Assemble: mix the kale with chickpeas and butternut squash pieces then mix in dried cranberries and add dressing as desired
Tips: Substitute your favorite winter squash (delicata, acorn, kabocha etc.) in place of the butternut squash or try sweet potatoes or even just regular baby Yukon golds instead. If massaging the kale seems too finicky or you are in a rush, use baby spinach or arugula instead – they can simply be mixed in without any extra preparation. Add slices of a tart apple for extra fall flavor or some baked tofu or tempeh for extra protein. In the latter case, you can marinate both tempeh and tofu in some of the salad dressing, perhaps with a little bit of extra oil, before baking them (my go to is 30 - 40 minutes at 400F with a flip halfway through). Steamed green beans or roasted brussels sprouts (just roast them with the squash) would be great additions as well. For an extra kick add some very thinly sliced red onion (about a third would likely suffice) or shallots (one or two).
As you are bound to have extra dressing use it on other greens or on top of roasted vegetables (maybe with a side of rice for a whole other meal).
If you have never had nutritional yeast rest assured that it is not alive and simply adds a bit more ‘funk’ and latent cheesiness to vegan dishes. I have found it nothing but salty when bought from spice companies but quite a welcome addition when more fresh in the bulk bins of my local food cooperative.