Computer Says Maybe

Alix: [00:00:00] Hey there. Welcome to Computer. Says maybe this is your host, Alix Dunn, and in this conversation I sat down with someone who has been studying for quite a long time how to make computer programming languages that are inclusive when she first reached out, because she listens to the podcast and wanted to share more about her research, which by the way, if you ever wonder if you should do that, yes, you totally should.

Alix: We love hearing from you all, and sometimes we end up interviewing guests. Who reach out cold. So please do if you're thinking about it. But she is really interested in how the experience of learning in computer science is really gendered and how that shapes who ends up being involved in building the technology that runs a lot of our lives.

Alix: She also makes. Programming languages herself. And I wanna just dive right into this one 'cause we get into a lot of topics that I thought I knew about, and actually I realized I didn't know that much and learned loads in this discussion. [00:01:00] So with that, let's get into it with a self-described involuntary ethnographer of computer science.

Felienne: Hello everyone. My name is Felienne Hermaans. I'm a professor of computer science education at Re Universite Amsterdam, and I research inclusion in programming language design. Sometimes I call myself an involuntary ethnographer of computer science because I. I made a programming language called Hedi that is inclusive, inclusive of gender and inclusive of different cultures.

Felienne: But then I realized that computing is not so excited about more inclusion, at least not everyone. And now my main focus is studying how did this culture of computer science came to be? What factors contribute to it?

Alix: I'm not even quite sure where to start with this conversation, but I feel like this idea of the technology spaces being masculine is one that is really common, but I think it's looked at in a different ways than the way that you look [00:02:00] at it.

Alix: The first thing I think of when I was reading a little bit more about your work is like James Damour and that period in Google where it was like, uh, men are just naturally better at technology. Yes. What is, you know. It's not our fault that there's a imbalance of genders in the organization. It's just, you know, yeah.

Alix: But the girls just don't want to do this. They just don't want, they wanna take care of their families, you know? Um, yeah. And I'm also

Felienne: thinking of Paul Grham of a Y Combinator that said one time, like, we can't make girls look at technology. Right. They're not interested. We can't force them to look at programming or technology or apps.

Felienne: It's

Alix: a really a historical yes. Way of thinking about this. Like I think about Mar Hicks and their book Programmed Inequality and the history of technology is full of women making huge breakthroughs and innovating in incredible ways. And then in like the eighties, nineties, when it started to become a profitable career, they were like.

Alix: Pushed out essentially. Tell me where do you think we should start?

Felienne: Let's go a [00:03:00] little bit back. Right. So the first start of computing was mostly female. I think most people know that that programmers or computers used to be women. And then the people programming the computers were women. So the first programmers were only women, and then it was sort of socially constructed as secretarial work.

Felienne: The men were doing their thinking and the women were just typing because of course that's an easy skill. And then after a while, people figured out like, Hey, there's something here. This is enjoyable. And I think the profitable even comes later. I think the first germ of wanting to own the technology space was like, oh, this is fun.

Felienne: Right? And then you see. Programming being made masculine. And one process by which this happens, there's a historian of technology called Nathan Ans Manger, and he wrote about this, how programming was made masculine by making it mathematical. So it was mathematized. And I think that process came before,

Alix: what was it, [00:04:00] before it was made mathematical.

Felienne: It was sort of nothing. And that's interesting. As a field, it didn't really have a very strong disciplinary identity, so there was a lot of electrical engineering involved because people were making computers, but many different people were involved. You might also think of a Noam Chomsky. A linguist was involved in early computing, especially in academic space.

Felienne: There was Herbert Simon, he's a political scientist, but he won a Nobel Prize for Economy and also a tour awards, which is like the Nobel Prize for computer science. So. A lot of different people participated in this field from different backgrounds because there wasn't a field, and that also meant that no one could stop people really from participating.

Felienne: This is what S Manger also says that early computing was especially open to women because it didn't have a disciplinary boundary. And what that means is the man couldn't say, Hey, but you don't have a degree in computer science. You are not welcome here because there wasn't a degree in computer science.

Felienne: So then you have this sort of a more field. [00:05:00] Which is coming partly from industry where computers are being made. Partly it's in academia, people are using the computers, and then how academia works is a field has to land in something like a faculty. So this field of computer science lands in academia and it has to go in a faculty because that's just how universities work and how they work.

Felienne: Then as well. And then math came along, or the natural sciences more generically, but specifically maths and math says, haha, I would like to have a piece of this pie. Because it was also coming from industry. It had money, it was in fashion because even then they were saying like computer cell algorithms are going to change the world.

Felienne: And then there's this one writing, I talk about this a lot in the mid seventies. He's a famous Dutch computer scientist. He says in the mid seventies. I want to present programming as a mathematical activity. So that's very late, right? This is 30 years after there were already computers. People were making programming languages for 20 years in a mathematical way, apparently.

Felienne: And then [00:06:00] he says, of course, it isn't only his fault, right? But he sort of embodies this stream of math people that say we would like to have computing. And from then on computer sciences in most places, indeed lens. Either in engineering or in the natural sciences, and that is the point. In which women are also driven out.

Alix: So basically it's not that it wasn't mathematical before, it's that it wasn't mathematical at the exclusion of other disciplinary fields and perspectives. So then it exactly,

Felienne: it was math, but it was also engineering and it was also linguistics and it was also cognitive science. So interesting. 'cause

Alix: now I feel like we're in this hand waving moment of saying, why is it only math people that have been involved in computer science?

Alix: Because it's causing all kinds of social problems because they don't have the three dimensional view of what it is they're working on because they don't know how to engage with other disciplines. So basically we had it figured out from the beginning, then stripped that multidisciplinarity out of it and now it caused lots of problems.

Felienne: And you also see that in the [00:07:00] history, like a lot of people. I also said this for a very long time. Oh yeah. Computer shines. Comes from math, right? Because in an academic setting, in universities, computer science, first word, where they were actually part of the math department, and then they split off into their own departments.

Felienne: So it's not entirely a lie that computer science came from math, but it's like first they took it and then they spun it off, but they still, of course. A computer scientist inherited this mathematical way of thinking. And indeed, exactly what you're saying. At the detriment of other forms of thinking, like, okay, so we can prove that this algorithm always work, but do people like it?

Felienne: Does it cause harm? Is it bad for the planet? That type of thinking absolutely doesn't fit the brain of computer scientists.

Alix: Yeah. And they probably want to keep hold of such an economically vibrant part. You know, academia now, like essentially that is the place to get funding, I assume, in an environment where university and research funding is collapsing.

Alix: Yes. Um, they're not gonna let go of it [00:08:00] easily.

Felienne: No. And it's still of course, also. A sort of status that you need. If you wanna talk about algorithms, if you wanna give talks, even outside of academia, having a PhD in software engineering, having a PhD in computer science that is qualifying you to participate in debate.

Felienne: So what I, what I see now a lot is that there's amazing philosophers and theologists and sociologists, and they also want to talk about computing. But then computer people say yes. But what do you know about programming? Have you ever made an algorithm? And I think, well, I don't care. Right? They have good new other perspectives to add.

Felienne: I'm still interested, even if they did never build an algorithm, because that's not like the holy grail of knowledge. So now this sort of the opposite of what S manger was saying is now true. Early computer was unusually open to women. To other ways of thinking and to other disciplines, and now it is entirely closed because if you are super [00:09:00] smart and you have a lot of good things to say, I am thinking of Emily Bender, for example, who is a linguist and she has amazing things to add to the debate on ai.

Felienne: And people say yes, but she's a linguist.

Alix: When I think of accessible programming languages, I think of Scratch, which I feel like maybe it's accessible almost like at an age level, like it's meant to be for kids. It's not seen as, as something that adults would pursue. But do you wanna tell us a little bit about the landscape of other accessible programming languages that have been built and the impact that's had and kind of who has been working on them?

Felienne: Yeah, absolutely. So. Scratch is a very interesting example, uh, because it's, it's not so accessible in many different dimensions that people might not know if they don't know so much about accessible computing. One thing is that this drag and drop interface, it's really nice if you can use a mouse, but if you wanna use a keyboard and a lot of visually impaired people, they do not use mouse because you have to see where the pointer is.

Felienne: It doesn't really support a keyboard interface very well. That's one form of [00:10:00] accessibility that's missing there. Another one is. Intercultural accessibility. Scratch. You can localize it. So you can, for example, send it to Chinese or Arabic, but if you actually do it, I have had some work on Arabic programming language design.

Felienne: It doesn't really work. For example, you cannot input Arabic numbers. Uh, they have different numerals, they cannot go in. And one, maybe, maybe the interesting, very, very invisible way in which scratch is not so very accessible has to do with gender differences. So in principle, scratch is not gendered. It doesn't have like things you associate with boys, like, I don't know.

Felienne: Basketball or robots or blue stuff. However, we know from research that boys and girls tend to go about learning things very differently in the technology space. Men and boys really like to explore, like to try out stuff, and girls and women are much more likely to want to know what the goal is and then have achieved the goal and then get some feedback that they've achieved.

Felienne: The goal. And this might have something fascinating. Yes. Uh, there's lovely [00:11:00] work by, uh, Mark Burnett. She has a framework called Gender Mag, and she looks at exactly this type of assessment and this maybe has something to do with it. Uh, girls and women tend to feel more insecure in the technology space, so maybe they're like, eh, I don't know if I know how to do this.

Felienne: And then maybe if they get a little assignment and they get positive feedback, then this might help them out of their insecurity. When I read that and then I looked at scratch again, I'm like. This is a problem because scratch assumes that you're going to want to figure it out. It doesn't have build in lesson plans.

Felienne: They have a few examples like, oh, here's three blocks you can put together. Everything is there, so immediately you have like 25 options or maybe 50 options of different blocks. So if you're the type of person, and this is much more likely a girl, but also kids from lower socioeconomic backgrounds and for the type of kids, it's like, I think programming might be really tricky.

Felienne: Let's see what this is. And then you open scratch. Now what? Right. You're less likely to have prior knowledge if you're female, and then the system doesn't really [00:12:00] take you by the hands, and that dimension of inclusion is really missing from that and also many other programming environments. This for me was like.

Felienne: A key knowledge that unlocked seeing so much in accessibility. I'm like, yes. Like also if you look at programming languages for adult users, so not scratch, but anything. Yeah. What do programming languages come with Ara, right? They come with Arapo, so you can immediately try stuff out. And a lot of early tutorials are like, oh, open arapo, which is like this.

Felienne: I find it still an intimidating text interface. A hundred percent. And you type something and then you get error messages. Yeah. Imagine. If we would take this modality of learning into accounts where we know there's a large part of the population, more likely female part, but also including men. Margaret Burnett makes us really clear in the writing where you would open a programming language interface, an IDE, or a tutorial, and it would say, hello, what do you wanna build?

Felienne: Choose these three tutorials, and it would say, okay. We [00:13:00] will do this and then you do it. They would say, oh, good job. You printed something. Next step. Right? I can almost feel how boys and men would be like, eh, right. I don't want someone telling me what to do. And women would be like, ah, this thing is helping.

Felienne: So I think that type of inclusivity. There's the assumption that you want to do programming and there's the assumption you want to figure it out. It's baked into so much of technology.

Alix: That's so interesting. It's true. Like I, like I was thinking about like the first time I used a GitHub repo 15, 10, 12, I don't know how long.

Alix: A long time ago. I remember how. Frustrating. I found it. And like the achievement or the reward of having completed it felt very disconnected from what I was actually trying to accomplish. It was almost like I wanted to pat myself on the back for like making it through the 19 steps and like being able to like read with specificity that one phrase that I misunderstood and then made a mistake.

Alix: Like it's such a punishing process and the like, the, the feeling of, [00:14:00] I dunno, I hadn't really thought about this. Until you're describing it in that way, but like there's almost like a top down like militaristic approach to it where it's like you have to feel pain during this process to actually come out of the other side with knowledge.

Alix: Um, yes. Which I don't. Like

Felienne: I also, yes. Yeah. Yeah. And sometimes people tell me this directly, right? So I have this programming interfaces called Hadi that does this, taking by the hands and many other things to make it more inclusive. And then I give a token and people, men, they come to me and they say, you make it so easy.

Felienne: You make it too easy. This way everyone can do it. I'm like. Yeah. Yeah, that, that's the goal, friend. Yeah. That is literally my goal.

Alix: But it's hazing. It's basically they experience this punishing thing, and so they're like, for you to be in the club, you also have to be punished.

Felienne: Yes. And so, I mean, some people enjoy this type of puzzle or they have this self reward, right.

Felienne: That they pat themselves on the back after the 90 steps, like, [00:15:00] oh yeah. I get it, but if you're not so motivated, if you're more goal oriented, then means oriented, then also this doesn't really fit. I know a lot of people, men and women, that maybe they wanna learn programming, but they have a very, very specific goal, right?

Felienne: They wanna make, I don't know, an app to track when they went to feed their horse. I don't know, just something they need for themselves. But then. The only way of learning is to help people on that path and, but then programming is never about goals. A lot of programming education, they have these assignments like I'm sure if you took a computer science course, you know these assignments.

Felienne: Here's a string and you have to find the longest alphabetical substring in this string. Let's print the first a hundred prime numbers. That doesn't tell you. Anything about what programming is for, that's only fun if you want to learn programming, if you don't care what you're building, it's never like, oh, the first thing we're building is something you might care about.

Felienne: Choose your own topic and we will together build something for that. [00:16:00] No, it assumes you care about programming already.

Alix: I took like econometrics in university and I remember that you always had to have this like breakthrough moment that was preceded by beating your head against a wall for like days and that there would be a moment and you're like.

Alix: Oh, I get it. And like that moment feels great, but I can totally imagine there being a type of person that's willing to traverse that feeling of uncertainty, stress, frustration, disorientation to get to that point. And you really have to like, I don't know. Yes. Had you hadn't thought about it. There's so

Felienne: many ways of thinking that are excluded or ways of thinking.

Felienne: If you think that way, you will not feel very. At home in computing. And so this, what you're describing is very bottom up learning. So first you get all the individual pieces and you have to trust that there's going to be a cake later. But now first you have to mix the egg and see the flour and trust that we're going to do something fun.

Felienne: But there's also people who are much more interested in, oh, here's seven cakes. Let's sample them. Which [00:17:00] one tastes best? Okay, now we're going to make that one. So this top down focus on the goal. If you are a person like this, you don't feel that, or here's another way of thinking. I have a friend and she's a journalist, and she told me recently that she took a programming course in university and then the professor said, okay, this is Python and we're gonna do print, and then, uh, round brackets and then quotes.

Felienne: And she said, why? Because she's an investigative person. This is why she's an awesome journalist. Yeah. She's like, why? And then the professor said. Don't worry about why trust the process. You just have to do this. Right? Yeah. There's also this, this military thinking that you refer to, and she was like, I didn't feel like I belonged there because I'm a person that wants to understand why I'm doing things.

Felienne: And I was like, I almost felt attacked as a person that has successfully completed many of those programming courses. I was like. Wow. I never asked this, right? I am, apparently, I'm not the type of person to do this. I was the type of person that like, okay, this is the rule. I will follow the rule because I enjoy programming.

Felienne: And I [00:18:00] thought this is also a modality of thinking that we don't support. If you are a person that's like, I want to know why stuff is the way it is, you cannot succeed in an introductory programming course because it'll force away upon thinking on you. And I thought, well, I mean, you can imagine with some.

Felienne: Efforts. But you can also imagine a programming course in which people design their own programming language. They get to decide how it works, and then from that you do some sort of programming. And this will be hard, right? It'll be technically hard, it'll be impractical if everyone has their own programming language.

Felienne: But I thought, oh, there's so many ways of thinking that's programming education, either discourages or the people that stay in like me. Sort of gets, somehow their brain gets smaller or the ways of thinking that they use gets limited, and other people, they just drop out. They're like, well, I'm a person that likes questions.

Felienne: If questions don't fit in this course, I will pick another course.

Alix: Yeah. You're also making me realize [00:19:00] that it's the same logic structure of venture backed company development, so like. Basically, it's a bottom up approach. It's like the only goal orientation is to make a bajillion dollar company, and you can pivot, pivot, pivot.

Alix: And the building blocks are things like product market fit, user testing, split scaling, like they try and isolate. These skillset as though they're content and goal agnostic. They're process goals. It's like if you do these things and you get good at these particular components, eventually you will reach this obscure goal and you might end up making a laundry delivery company instead of, I don't know, like an educational tool.

Alix: But that doesn't really matter because you've essentially achieved the core goal. So it's like it's yes. 'cause you're an

Felienne: entrepreneur as identity. This is. This is a good point. It's, it's always weird to me that people are like, yeah, I wanna, I wanna found a company. It's like, to do [00:20:00] what? To do what exactly.

Felienne: To, to make what, and then they don't care. And this, I, I had never made this connection, but I think you're very right. It's a similar way of thinking that I don't care about what we're building, what I care about is that we're building.

Alix: Yeah, there's a punishment as culture way of thinking within venture-backed companies

Felienne: and also, uh, a non-commitment to the goal, as you're saying with pivoting, right?

Felienne: Oh no, I make this, oh no, I make something different. By definition, you cannot have an attachment to the goal because then you'll be sad. No, you'll be sad to pivot if you really like the first goal. So that probably also creates a culture in which a goal you aren't. So. Spiritually attached to it because that would destroy you.

Alix: Totally. And the only goal orientation is money, prestige and scale, basically. Okay. So do you wanna get specific about languages you've built? Like how does it work? What do you think are the most accessible languages? Like what has the uptake been like? Say more about specific languages.

Felienne: So maybe let's do one other language first that isn't my own language.

Felienne: One language I think that really has accessibility nailed down is [00:21:00] elm. So this is a, a browser based or a browser oriented functional language, and they've really, really thought of how to do error messages. And the creator of ELM really said, I don't want beginners to drop out. I want beginners to stay in.

Felienne: So he like manually reads, error messages and reads where people get stuck and you could just see it from their error messages. That's. Care and love have gone into it. You then see this in a community that that's also more inclusive. I mean, a creator also sets a culture, but the language also, I think it just tells you, you are welcome here.

Felienne: It doesn't matter that you made a mistake. I will help you fix it, rather than something like invalid character, right or or unexpected endo file while parsing. That's not helpful.

Alix: I also really love their design and their website. Like it's playful, like the implication from their design, just looking at it quickly is that they're interested in play.

Alix: Whereas I feel like I'm thinking of a corollary example that tries to onboard people. You [00:22:00] get confronted pretty quickly with like pricing opportunities and possibilities once you become a programmer and like all this very. Functional information that I'm like, Ugh. I guess I go into the grinder and like learn all this stuff for a purpose that, as we've described, is unclear at this point.

Alix: But I'll be powerful and worth money at the end.

Felienne: Yeah, and I think Elm is really inviting and then it's also follows true on the invitation. So it's like, come here and then they keep you in. So I really like that.

Alix: I mean, I'm curious if there's all these new languages. If they are like tributaries into a river of more common programming languages or like how, if I'm onboarded and I get really into it after being welcomed in a way to thinking about programming and then I reach the end of Elm as a language.

Alix: I imagine if I were to apply for a job and I were to say, the main language I can write in is Elm, they'd be like, you don't? Yeah. Yeah. So like how does that work? How does that work to like connect these types of experiences to bigger languages?

Felienne: Yeah. And so this is also really a problem [00:23:00] with our, my own platform.

Felienne: So my programming language called Hadi, and it's very much made for education. So it's like a baby python that slowly brings you to Python. And one of the things we do is localization. We, we do that very well, I think. So you could program in English, but also in 69 other languages, including Dutch, but also Chinese, Korean, Japanese, Arabic, Russian, Ukrainian, many, many languages.

Felienne: And initially I thought. Because I'm a technologist. Well, maybe I'm a former, I'm a recovering technologist. Let's say that I thought if I build Hadi, which is basically Python, if I build Python like in Arabic, then people will see how to do it right? And I wrote a lot about, this is how you do it. I thought I'll show people how to do it and then they will do it.

Felienne: But then now I'm seeing this is also why I'm now so much more interested in this culture of programming, that the reason that it didn't exist is not a coincidence, right? It could have existed in theory in a different community, but these nice languages like Elm or Hadi, [00:24:00] that's shield you from some of the issues, they don't perpetuate.

Felienne: Because of the culture of big programming languages, and if we want to have this, then first the culture must change, which is super, super hard. I thought it was hard building an Arabic version of Python, but that's not so hard as changing a culture. So still many, many things don't work as a part of this experiment.

Felienne: I tried just to add one plus one in Arabic. They have a different character. For one, we have like a one stick and they have like a, an a skewed stick a little bit to, so if you just type wahaha, so this is one plus one in Arabic in any of the top 10 TBI index programming languages, they all crash. They all give an error message, and the error messages are sometimes invalid character.

Felienne: And I'm like, what? Right. What do you mean invalid character? This is a one for 300 million people. What do you mean invalid character? You are invalid. This is invalid. You're

Alix: invalid. [00:25:00]

Felienne: So most languages, Python, also, they do support non KY characters. You can have unicodes, uh, variable names, for example, but the numerals.

Felienne: They're just, they aren't supported because people don't know about it because it's so very hard to penetrate this culture. A lot of people that I know that are programmers and Arabic speakers, they from the beginning on just. Agreed, accepted that, oh, we must do everything in English. So they accommodate and then they accommodate so well that they sort of forgets to ask about, Hey, but wouldn't it also be nice if our numbers are included?

Felienne: Sometimes if I show Arabic Hadi to people from an Arabic speaking country, they're like, wow, right. I never imagined I would see programming in Arabic. So in order to participate, you have to leave your culture at the doorstep, and then you get used to it. So sort of seeing this. I mean, I know this sounds a bit overly dramatic, but seeing all those programming languages fail to just add one and one, [00:26:00] it destroyed me or it destroyed part of me.

Felienne: Like how can it be that we have a culture of programming language designers and no one cares about this, and it's just numbers, right? There's so many other things that you might wanna fix, but it would be. It would be a one hour fix probably because those characters, nobody knows about the character, so they're not gonna mess up your barhop because they're not used anywhere else.

Felienne: So it would be so easy to fix. And then how is it not that people are like, I wanna fix this? From then on sort of, there's this break in my career that for a few years I was like, I will fix this from a technology perspective. And then this got burned into my eyes and then into my soul. And I thought my only goal in life now is.

Felienne: To study this culture, so I call myself sometimes an involuntary ethnographer of computer science. I just want to understand how we came to live in the world. It was, I cannot add one-on-one in Arabic, in Python. I just, I don't get it.

Alix: An involuntary. What

Felienne: [00:27:00] does that, say that again. So I describe myself now as an involuntary ethnographer.

Felienne: So someone that does ethno, I'm like studying the computer scientists. In the wild. Oh, what ritual are they executing? But I am a computer scientist, right? This makes it so sad. I love programming and I love building programming languages. I'm so proud that I can call myself a programming language designer, right?

Felienne: I made a programming language and hundreds of thousands of people use it in the whole wide world. It's awesome. And now, I mean, I used to think it was awesome, and I'm sort of ashamed that this puts me in a category of people that just. Have no interest in other cultures and have no desire to fix these problems.

Felienne: And as I said, this is why I was like, my goal, my research goal now is to understand from a historical and a sociological and an ETHNOGRAPHICAL perspective, what the fuck? Right? Why, why is this the way it is?

Alix: I [00:28:00] imagine it's only gonna get worse. So I feel like there was a period of time when, with the sort of context of Gamergate, where there was this kind of push and pull inside companies about the role that women, or not hyper masculine, like engineering nerds, um, like what role they should play.

Alix: It was also a time when these companies were seeking rare talent and I feel like there's been a. Shift because of the, both the size of these companies, the market saturation of like people that now know how to do programming. Also, the invention of technologies that purport to do that work on behalf of workers.

Alix: You don't, you don't need an employee anymore. You can just use some broken, um, large language model to help you with your, with your coding. But that now it's like the tech companies are not. Trying to increase the number of people that have these skills and expand the diversity of the pool of candidates and the kind of [00:29:00] cultural diversity of the people that are interested in programming and, and working in technology industries.

Alix: It feels like there's even less of an incentive for these changes to be made because of how like the economics of technology is changing. But I don't know if you're, do you feel like there'll be like. More or less demand for creating these pathways into learning these systems.

Felienne: Yeah, so definitely all the things you said, and I think I wanna add one thing to it that a lot of inclusion, I think also.

Felienne: Wasn't really like initiatives for more women in technology. They weren't really coming from the top, they were coming from employees, but because there was a demand for programmers, the top wanted to keep programmers happy. It's like, oh, if you wanna have your women in programming circle or black people in tech circle, sure.

Felienne: Right. If this is what's going to excite you to work at our company. Then by all means. So I think there's this extra dimension that I, I don't believe Mark Zuckerberg ever really cared about women in technology. So the [00:30:00] fact that he's now saying stuff like, oh, technology has become too feminine, right?

Felienne: This is like one of the rigid guys on the planet saying, oh, technology is too female. He always believed this. I don't doubt that he always believed this, but he was afraid to say it because his workforce would leave and he cared. He hold on himself, his. First

Alix: product was hot or not like, I don't know. I like how did, how did we ever let that 10 year period where they pretended to care about the diversity of their workforce happen?

Alix: Like it's the stars aligned with a cultural movement and a moment and a political moment where they had to like pretend, I completely agree with you. Like he was never truly invested. In any of this stuff.

Felienne: And then now indeed, these e economic conditions make it so that he can just say what he wants and then if you don't like it, you can not go work for another company because no one is hired.

Felienne: And I think to come back to your actual question is yes, I do think it's harder in the us There's a lot of. My fellow researchers doing inclusion in the digital space that are having their grants revoked, [00:31:00] or they have to write their grants in such a way that it is about accessibility, for example, for blind and deaf people, because that's in the law.

Felienne: This is still research that can be done because it's not canceled. But if you wanna do race or gender or critical theory in the digital world or, or in any world, but let's. Stick with the digital world, then you're just not gonna get any funding. The funding you have will be revoked. So yes, it will be much more difficult in the academic space to do this research.

Felienne: And also I think just in the vibe, uh, many men I think in technology are to a certain extent, mark Zuckerberg, that they also never really, really, really cared, but they sort of had to play along. So if you wanna submit a talk to a conference on the, I dunno, feminism or programming language design, then maybe your chances now are also lower.

Felienne: Because conferences will not be like, oh, we have to care about this. And of course there are many awesome people in technology, men and women that do care about inclusion. But it will be harder for them to say, oh, this looks like a good talk. Yes. Let's talk about [00:32:00] bias in, in algorithms. Let's talk about racism and sexism in technology because there will be other people.

Felienne: Who are now not so scared anymore to force their actual opinion, which is that none of this is necessary because women don't want it anyway. Right. They're not interested in technology. Citra.

Alix: Yeah. Because women aren't interested in fulfilling careers or they can make a lot of money. Yeah. That's such a turn.

Alix: Independence. Who

Felienne: cares? Yeah.

Alix: Um, okay. So you think basically the kind of political tension of this, making it accessible versus a punishing path that is male dominated is just gonna continue to exist and show up in different ways?

Felienne: I mean, I do believe ultimately stuff will get better. Everything goes like this and then it goes all the way the other way.

Felienne: That's true. And pendulum swings back. Probably over time. The arc of justice will bend the right way, but. The time we are in now is looking quite bleak, especially in US and US influences of course, how other people exist as well. [00:33:00]

Alix: Yeah, that's interesting. So talking about industry and their interest and willingness to hire people based on their experiences and like attitude about technology, that's kind of like the most downstream, I guess, manifestation of these issues.

Alix: If we go all the way upstream and we think about like. Youth education or like in school or in university environments, how these technologies are maybe making the path towards degrees in computer science more accessible for more people, creating opportunities for people to get excited, to get enough of an understanding of what they would be learning to actually know if it's something they wanna explore.

Alix: Like how are you seeing that space evolve? Like do you imagine over the next 10 years there's gonna be a lot more women graduating with computer science degrees? I actually don't know the stats on these kinds of things. Or do you think computer science as a field I. Because it's structurally not grappling with these issues is never gonna be the path, and it's always gonna be kind of smaller, more bespoke opportunities that are really accessibility oriented that people are gonna seek out.

Alix: Like how, how is that, how is like traditional education gonna deal with these [00:34:00] issues, do you think?

Felienne: Yeah, so I think to a certain extent, it's also this hazing that we were talking about before, that I have seen initiatives in different places in my own university and other universities. Where people, for example, try to change introductory programming from C or c plus plus into Python, which I think is a step forward towards accessibility in a certain meaning of the world, because it's just easier, right?

Felienne: We can just say without much controversy that Python is easier, less furnishing than C or c plus plus. And then I've seen faculty fight this. This was the hill they were willing to die on because it'll be made easier. Take home with ridiculous arguments, right? It's like, yes, but students, they have to learn memory management.

Felienne: I'm like, yes dude. I'm not saying they should never learn about memory management, but should it be week one, right? Should it be in a programming course that's about programming? If this is important, we can maybe do it in, I dunno, semester two or three. As I said, I am a recovering technologist. I also felt in, earlier in my career, if I make an easy [00:35:00] programming language, stuff will be fixed because then people can use it and then it will be easier.

Felienne: But the field as it is now is made of people who either because that's just the way they are or that's the way they've become by repeated exposure to computer science culture, they like hard stuff. This is a lot on my paper, is the case for feminism and programming language design. What it is about, they like getting status from doing hard things.

Felienne: In my paper, I make a comparison to the field of glaciology, which is the study of glaciers. This paper shows that. Harder to reach. Glaciers like high in a mountain in Antarctica or something are much more studied than rural area glaciers where it's very easy to get data because from this high mountain, you can get status, you can get bragging rights, you can climb a high mountain, and then you take ice there, you go back and then you say, oh guys, you know what I did today?

Felienne: I climbed High Mountain. And then all your friends are like, yeah, bro. And I read that and it's not about, and there's nothing to do with programming. But I read that [00:36:00] paper, I was like. This is us, right? You can gain much more status from being able to do c plus plus close to the metal, right? Look at the metaphors that people use for this.

Felienne: Close to the metals, like, dude, are you touching a motherboard? No, it has nothing to do with metal, but this is much more status heavy than. Python, right? And then you do it maybe in an environment that you do data science, maybe it's, it's a much interesting, much more goal oriented environment if instead of doing Hello World or printing prime numbers, day one is here's a, a climate data set or a data set from a library or something you might actually care about.

Felienne: And we're gonna analyze this data set with Python, but. I think it's easy for people that are exist in this technology space to see how doing a dataset analysis of something that is interesting. It's very, very different from printing some sort of inter sequence in c plus plus. So purely the existence of easier languages is.

Felienne: Not [00:37:00] helpful or at least not helpful enough. How do we get outta that?

Alix: I mean, I feel like, I mean, your point earlier about like you were a technologist and then you were like, oh, actually the problem is culture. What does one do about about some of these core cultural barriers to making all this work better?

Felienne: I don't know. So one thing I did, I'm not sure if it's helpful, but it was helpful to me. It was a bomb for my soul, is to pick this culture apart. So at one point. Three years ago, I was just so fed up. It's like. Why do I feel so bad if I hang out with these people? Right? I go to a programming languages club, conference meeting, whatever.

Felienne: I'm a programming languages person. I go there and I come home sad and depressed, like why? And I wrote this whole 20 page expose on like this glaciology stuff. It was helpful for me. It was like therapy almost. It was like processing, but what I'm seeing now also is this, this way of thinking. It is helpful for other people since I started to do this [00:38:00] work, so many people, especially young women or young non-binary people, they come to me and they say, now that I've read your work on this, hard as valuable.

Felienne: Now I understand my own feelings pattern. Now I have a vocabulary to talk about these problems. So that's not a fix, but it is maybe a necessary step where I know so many people now in their vocabulary about computer science is the term hard, is valuable. And they say this to each other, right? People say this to me as well.

Felienne: We go to a conference and there's a guy again with 25 Greek letters and a proof, and it's something, something to do with programming. And the whole room is like. We also don't know, and then people will say, ah, there's a hardest valuable guy again. I do think understanding of picking apart the culture is one thing that needs to be done well.

Felienne: You

Alix: make it explicit. You basically say like, there's this thing happening that we've kind of accepted as a norm, but it's actually a very specific way of thinking about things and there's a lot of other ways to think about things. Um, I also wonder if [00:39:00] like the heart is valuable contesting, it is partly about disambiguating, no code energy from accessible programming education.

Alix: Because those are two very different things and I feel like we're in this era where like anyone should be able to use no-code to like make whatever they want. Like use a chat GPT to make an app. You don't have to understand how anything works. And I feel like there's probably a lot of technologists right now that are also feeling threatened by the number of people that are basically loudly proclaiming excitedly that they don't need them anymore.

Alix: Distinguishing between accessible. Programming experience that allows you to better learn, you know how to be a programmer versus basically saying you don't even need to learn how to program to do these things. There's these like superficial technologies that don't worry about what's under the hood.

Alix: We'll like make it drag and drop essentially, and you can build like whatever you want. And I feel like there must be, I don't know, are you seeing your work [00:40:00] cast in a different way now that there's more like emphasis on no-code technology and building.

Felienne: I do see my work cast in a different light. Now there's just more LLM programming.

Felienne: People will say even more like, oh, you don't need to learn programming because you can use chat GBT. I'm like, well, but then you'll not learn anything, right? You cannot see what it's doing. And also you'll need to be on an app all the time. That is maybe not free forever. So I don't see this as necessarily the democratizing or equalizing the playing field because some people might know how to program after a while and then they don't need an LLM.

Felienne: And some people might be bound to expect expensive tools. I recently made like a little app with LLMs and I was. Sort of surprised about. I mean, it makes many mistakes, but then after a while it makes, it makes a nice CSS that I'm not really good at. And I thought, oh, this is useful. But then I also thought, okay, accepting that this is useful means that I will never learn CSS and maybe that's okay, right?

Felienne: Maybe I can live with this because I [00:41:00] know many other parts of technology. It, it's also giving in. It is accepting that this is too hard for me and that I can never learn. I would be sad if people say. Have that relationship with all of programming. Right. It's much different for me to accept that I cannot do CSS, I'm still a programmer, but if many people are like, oh, I never need to learn this, then I don't know if, yeah.

Alix: Then what happens. Yeah.

Felienne: Yes, and I don't know. I mean, maybe it's fine, but I wouldn't describe it as democratizing because it's still, the actual knowledge is in the hand of a very small set of people, of which many people do not share how I look at the world.

Alix: Yeah. You're also exposing yourself to the.

Alix: Likely abuses of platform power. It's one of the things that has always had me return to trying to understand what's actually happening. So I found myself watching like a 90 minute video of a Stanford professor trying to explain how to train your own LLM, and I found myself like downloading all of the requisite programming environments and all these things, and like I found myself doing it because [00:42:00] I find it so.

Alix: Dangerous when like all we have is a prepackaged set of components that are offered up to us as consumers. And it feels really like the further you allow yourself to be from the actual. How the sausage is made, the more likely you are to be taken advantage of, priced out, intimidated. If you go too long, you actually can't kind of reverse engineer or teach yourself enough to kind of know how the constituent components fit together.

Alix: Because technology changes so quickly that you can be really quickly out of your depths, which maybe is actually part of this conversation too, that like even hearing some of the language that's used in deep learning, it takes ages to like look up what they're talking about and like how it all connects and how what these technologies, how it all fits together.

Alix: Academia, you're supposed to stand on the shoulder of giants, and sometimes I feel like I'm standing on the shoulders of really avaricious tech exclusive [00:43:00] money seeking power seeking industries instead of generous people that want to teach me how it all works.

Felienne: Coming back to these modes of thinking, when I was making this little app, uh, with, uh, with lms, then there was one thing I didn't want.

Felienne: So I couldn't upload a PDF that I was making a PDF, and I wanted to have a random page of the PDF so that I can look at my diary and I get a random page every day. I thought this is a cute idea. But then it didn't let me upload over a hundred megabytes, and it took me a lot of effort to actually figure out why.

Felienne: Because of course this was in the sort of back end, but also in the front end to get an error message. And there were, there were ifs on both sides. Wasn't necessary I think, but that was the way it is and I found myself. Almost thinking, okay, nevermind. Right? I'll just make my PDF smaller. I'll just search for a tool that compresses my PDF so that I can accept that this is the way it is.

Felienne: And of course, I could do, I knew I could do it, so I did it, but I thought, oh, it's so easy [00:44:00] to get into this acceptance mode. So instead of being a creator of technology where I'm like, I have an idea and I will make this thing, do my idea, because that's. That is what I think programming is or any creation is.

Felienne: I have an idea and whatever technology I use. Most bend to my will as a creator, right? I will. I will mix the paints so that it makes the painting I want. And I just felt like this acceptance, like, okay, maybe it's fine as a this, I said, no, stop. Paramount. Stop, right, come back. But I can so see how people with less knowledge and less confident, right?

Felienne: I am so confident that I know how to do it. I have techniques. I know how to search codes, but I can so see how people with a little bit less skill, not even entirely novices, right? But even people, students that have maybe two or three years of programming experience, they would also go back, like they hang back in their seats.

Felienne: It's like, mm-hmm. Yeah, okay. A hundred megabytes also fine. That.

Alix: It's really

Felienne: worrying. I don't have a word for that. It's like [00:45:00] intellectual laziness. It's not exactly the same because it also has to do with the confidence of knowing that you can do it. But I thought this is what happens, that this the ultimate, this spark in you that you're like, oh, I wanna build it.

Felienne: It'll be. Small and lessened by technology that has its own will I say with big air quotes and then you just succumb to it. You're like, yeah, I guess that's also fine. No, it shouldn't be fine.

Alix: It shouldn't be fine. No, and I think, I think over time I really worry about how institutions will relate to the knowledge of what they've built technologically.

Alix: So like in that situation, like you might have someone with less experience trying to build systems, use some technique that allows for them to. Make this app that has this a hundred megabyte limit, um, and then that person doesn't necessarily know why that's the case. But then increasingly the knowledge of an institution, of the technology that it builds increasingly lies in the companies that build these platforms and within the models and [00:46:00] what gets produced by themselves.

Alix: And I think that that gives you less technical intuition. It means it's harder to fix things, it means it's harder to feel confident about technology. And over time it's like this new form of outsourcing. Technical thinking. You know, when you think about Google Workspace really easily, being able to set up email is like a game changing, time saving thing in the world.

Alix: I'm super supportive of a company, especially one that provides free email services. That's great. Like that completely changed how people use. Technology. 'cause I don't think that many people should have to understand MX records in DNS. I don't think that many people should have to like figure out, I don't know, like which email client is compatible, how to set up your SMTP server, whatever.

Alix: Like, I don't think people should have to know how to do that. And maybe there's a, there are people out there that do really believe that people should know how to manage their own servers as part of an increasingly lost art. But I, I do feel like we're increasingly cognitively offloading deeply strategic and functional aspects of how we want [00:47:00] to build things and maintain things.

Alix: And I think that to me, that's one of the biggest dangers of the LLM programming and the no-code space. It just outsources. Thinking.

Felienne: Yes. And also pleasure for lack of a better, yeah, so I'm, I'm learning Arabic because of this Arabic programming language. I'm now excited to learn Arabic. At first, I was on Duolingo, which is kind of expensive for what you get, but also it tells me how to learn.

Felienne: And now, and I did this without LLMs because I tried with LLMs to make a program that I could practice by Arabic, but it doesn't work because technology doesn't really work very well in Arabic. But now I made it myself. So I made my own little app Nice. And it like to practice and it gives me such joy, right?

Felienne: The, the joy that I made this, it's not even on GitHub. I try to put it on GitHub, but GitHub doesn't support Arabic names for repositories. Fun fact. So it's not even on gi Gi, oh my God, yes. Oh, it's so bad. If you wanna go to GitHub and you make a repository with an Arabic name, it doesn't do it. And the clients, or also the web [00:48:00] clients, it just forces the name to dashes.

Felienne: So you put that a few characters in and it just tells you, no, no, you don't exist. Here are four dashes. So this

Alix: huge world language is an error, basically. Yes,

Felienne: yes, yes. Yeah.

Alix: Great.

Felienne: It's, so then I was also like, okay, fine. I'm not gonna put it on GI off. I don't wanna, I don't want to, anyway, I. It gives me so much pleasure that I made this for me, right?

Felienne: That it allows me to practice very efficiently, be much more efficiently than Duolingo because I do exactly what I want in exactly the right way. But this joy of having built something, it's not even cognitive offloading. I don't know what it is. So offloading this proudness, and certainly with my random page diary app, I don't feel anything.

Felienne: I'm not even excited. To use it. Or I was excited about the ideas and now I'm, now I'm saying this, maybe I should also build it by hand.

Alix: Dot dot.

Felienne: Yes. And it's also hard if I just make it ugly, right? CSS is my [00:49:00] problem. I could just make a plain text HML page and I could easily do it, but it wouldn't look nice.

Felienne: But I think that's also so fundamental that if I teach young kids programming. We have a turtle also in in Haiti, so you can draw little drawings and then they make like a rainbow heart or something, and you just see a 12-year-old is mesmerized. But I think like Luke, I made a heart and it's a rainbow. I changed the color and they're just so proud and happy.

Felienne: I don't think. Going into Dolly and typing. I want a Rainbow Heart elicits feelings that are similar to programming a rainbow heart. I don't think those are the same feelings, and I like having those feelings and I like other people to have those feelings.

Alix: I'm with you. I still have that spirit too, like I can spend two hours doing something that I could have done in five minutes.

Alix: 90% of what I wanted. Yes. But if I do two hours, I get the a hundred percent. Yes. Exactly. And the sense of satisfaction every time I use it. Yeah. And I, I think that's really important. Some

Felienne: acceptance, like Yeah, it's [00:50:00] fine. Fine. It's not good enough. It should be fantastic.

Alix: Totally. Okay. Well, turning to kind of a concluding question, I'm wondering if someone wants to get involved in either.

Alix: Sort of experiencing one of these more, um, accessible programming languages or maybe making their own, if they're interested in it. Where should people start? Like where are some good entry points for people to, to get going?

Felienne: Yes. So I would like to say stuff I've done.

Alix: Please do.

Felienne: Hey, these open source, it is on GitHub and we are always looking for people that wanna help, wanna add translation.

Felienne: So you can go to hei.org/join and it explains to you how you can join as either a programmer or a translator. I wrote a paper. Three years ago about how to make programming languages localized. So if you work on a programming language you have, or Python or a small, a smaller one, and you are. Questioning.

Felienne: Could we localize this? Could we add Arabic numerals? Could we add gender accents? All these things that other natural languages might support, [00:51:00] want to have supported. You can check out my paper. It's called a Framework for the Localization of Programming Languages, so that will certainly help you from the localization perspective.

Alix: We'll definitely link to those in the show notes

Felienne: and, and just knowing about this and spreading the word right. Figure out how these things work and try them. Now that I know a tiny, tiny bit of Arabic, any software I see, I test it in Arabic and it's always bad and I always send screenshots to the people that create the technology.

Felienne: And I know they send me emails back saying, yes, if we have more time, we will do it. But this is, it is not super hard to learn the. Arabic alphabets and to know how to switch your keyboard setting or, or pick any, any other non Latin language. Learn a few characters and then any software that you encounter, you just try to input these characters.

Felienne: You don't even have to know what they mean. You just pick a random language. Any software you try, the characters, they will feel. It'll be an empty square. It'll be a question mark, be an error message. You send this screenshot. To the makers of technology. I know it's a small thing and you will be perceived [00:52:00] as an annoying person, but this is a thing you can do.

Felienne: And also, if you make software, then you can send yourself an an anno email or you can go about fixing it. So if you are creating any technology, I love this, try this, and you'll be, I love this. Surprised how shit, so be

Alix: a. Be a squeaky wheel for inclusion, um, and accessibility of every program language you encounter and just get used to that.

Alix: I love that as an idea because I feel like we're so

Felienne: smart, even it doesn't even have to be a programming language.

Alix: Uh, amazing. Okay. Well thank you so much for sharing all this. I'm so glad that there's work like yours happening and that there's this type of effort and also just that like. I don't know, perspective shifting, thinking about why have we gotten to a point where there's such limited multidisciplinary thinking in computer science, I think is just such a.

Alix: Refreshing take. We've been there before. We can be there again. And I feel like we should advocate for these spaces to be, um, more permeable to all kinds of understandings, expertise, perspectives, and people, and stop being so [00:53:00] exclusive in mathy about everything related to technology. So thank you. This is great.

Felienne: Thanks for having me.

Alix: Okay. I found that a little bit mind blowing when thinking about the history. Of computer science and how it has shaped who gets involved in it, and also just like the learning experiences people have and what we've come to expect and how that in and of itself can be gendered. Just super interesting stuff, and also just thinking about how stilted our digital spaces are and how much we might take for granted as just kind of the way things are.

Alix: That was an intentional process by which these languages got more and more exclusive, got more and more mathy and sciencey. And I've kind of stripped out a lot of the soul and creativity and energy and inclusivity that the computer science field started with. So a little bit of a, a different type of episode, but I think it really does touch directly on technology, politics and some of the undercurrents of misogyny [00:54:00] that cut through programming and related fields.

Alix: Thanks to Georgia Iacovou and Sarah Myles for putting the episode together, and next week we have our. Download from Fact where three of my colleagues, Soizic, Hanna, and Georgia, all were there on the ground reporting live, meeting lots of researchers, reading lots of papers, nerding out as I did last year in Rio.

Alix: Um, this year they were in Athens. Next week we're gonna hear from them what they learned, what stuck out, and hearing them describe certain findings from research that they thought was cool. So join us next week, uh, and we will dig into fact. And thanks for listening.