Hello and welcome to the coziest place in the internet where we will have conversations with computers. I hope you can all see me alright and all this open source and internet magic works perfectly. I will start to give you a bit of context and a little bit of an introduction of what Conversations with Computers is and how it came to be. I was working as part of the Servus Research Lab, which is every second year when there's not the amazing Art Meets Radical Openness Festival. And Servus is like this net culture initiative based in Linz, who is doing this work, running servers, running services for artists and cultural workers for 25 years. And together with them, we have developed the Silicon Friend Camp this year where we came together with 17 artists and explored machine intelligence together and as part of this framework of my own research together with Servus AT we kind of developed also this program additionally where we invited amazing guests who already did research on machine intelligence and computers and how this will be evolving in the future the Silicon Fan Camp did an exhibition which is called Camping with Computers. And this is still online on the page that you're viewing this now, but you can also go to Linz. It's in the WHA gallery. And yeah, first of all, I will start with a little thanksgiving because there are so many people involved in running this program my main gratitude goes out to David who is the co-organizer and also helped writing all the applications finding all the people and yeah making this whole program work second there's Christoph Nebel from the Zeitblasierte Medien, which is a study program in Linz, who are supporting Servus.at and the research lab since 2008. And of course our main sponsor, which is Linz Kultur and Linz Impulse in general and yeah there's also the Bundeskanzleramt and the Land Oberösterreich so it's like a very Austrian funding cultural initiatives second there are lots of there's the University of Art and Design involved with Silvia Leitner, the Gebäudentechnik and the Zentrale Informatikdienst, who are also part of running the servers now. We have Giacomo Piazzi and Antonio Zingaro, who are in the background doing all the streaming magic that you can see on your screen now. And additionally there's the Interface Cultures program that helped us with technical support. We also have some students who are silently running stuff in the background. There's Fatima Elkosht and Sara Pineros who are moderating the chat. So please be nice there. And yeah, finally, there's Manuela Nauvoo and Lasse Scherfik who helped me conceptually with setting up the symposium and getting like a nice mix of speakers together. So how do we structure this night and the upcoming night? It will be online only. We were initially planning to do this in person but unfortunately numbers are rising people don't get the vaccine as often as they could in Austria and so we have to all sit silently in front of our computers in our homes again because this is apparently how we do things now yeah so today and tomorrow night we will have short input presentations from guests who did already amazing research projects and researchers, designers and artists. And then we will have like a short conversation together the the speakers and me and i will try to take input from the audience who can participate in the etherpad so feel free to write down anecdotes or comments or questions and i will do my best to bring it up then in the end in the conversations. I hope this was not too fast sometimes I tend to speak too fast you can also send it in the chat that I should repeat some stuff or but in general we will try to yeah have like a fast-paced we will try to have like a fast-paced lectures or lecture series. The day to day will be about humans behind technology and emotional technology and we will have two speakers only here live and then there will be a recorded speaker, let's say. It's a bit of a surprise. And our first guest cannot join us, which is Lasse Scherfik, who's unfortunately sick. So all the best to him and maybe he's still able to watch even though he's not able to have a conversation with us so now instead of Lasse we have a student of his which is Konrad Weiser who's a Cologne and Klug-based artist and researcher and currently he's a diploma student at the KHM in Cologne and his he will talk about his current research projects which I found very very fascinating according to labor and automated systems and people behind yeah bots and other automated devices. So now I try to smoothly give my voice to Konrad, who will introduce his research project with the title, one second, Human-Driven Condition on the Relationship between Shiny between shiny interfaces and precarious work. Can you see the presentation? Nice. So hi everyone. So, hi everyone. Let's start with a little thanksgiving. It feels really nice to show my research within this context. And I'm very excited to have conversations with all of you. So today I will talk about my investigation to trace humans within computational systems and my ongoing research of automated engagement and bot-driven attention. In the 2011 book Human Computation, Louis von Aan and Edith Law first and foremost criticized the loose meaning of the term itself and want to create a clear distinction to the research fields such as crowdsourcing, social computation, socio-computational systems, or collective intelligence. Following their proposal of a first definition, human computation is simply computation that is carried out by humans. But the reappear comes with apparent in this second definition stating that human computation systems can be defined as intelligent systems that organize humans to carry out the process of computation. Whether it be performing the basic operations, taking charge of the control process itself, or even synthesizing the program itself. Yeah, here they're speaking actually about the system and not about humans, even though from the phrasing maybe it seems like it. Introducing the obsession of the machine's explicit control within human computation in the same named subchapter, and von Ahn leave no room for change by commenting on the related concepts of research fields that none of the related concepts emphasize the idea of explicit control. A human computation system differs their crowd-driven competitors like Wikipedia in the level of explicit control. While systems like Wikipedia do enforce rules, protocols and standards, the level of control of the laboring individual is still too high for law and fun. The field of human computation systems are magical wonderlands to digital tailors, having a global precarious human resource a few clicks away to deploy their next productivity, efficiency ideals and exploitative fantasies. These exploitative fantasies come together when Fan'an wants to reuse wasted human cycles for his capture endeavour, what he also calls the largest distributed human collaboration project in history yeah here for example you can see that he's actually dividing the the day the hours of a day and two seconds of the day are reserved for solving captures which then of course resulted into google selling about half a million hours of a day of uncompensated human labor. Within this realm of human computation, the whole automation discourse with all its ideologies and social political implications of programmable, addressable humans is reduced to a single line of code. The line of code and the comment which also gave the name for this project. The narrative of automation's technocratic workings hides the human and the labor behind a precisely positioned technological determinism and tries to draw them obsolete. It actively challenges the value and agency of work and reinforces imperatives of control. Automation became an ideology, and with it, its surveillance and precarious structure swept into every so polished interface. Throughout this exploration and in search of the people within, I encountered the phenomenon of group control systems and decided to define this technological arrangement and its cultural context as my research object to further investigate a specific constellation in the realm of human computation and automated engagement. Peopled Codespaces In mid-2019, first imagery of group control system appeared on social media platforms like Twitter. Filled industrial halls and tightly packed rooms with arrays of mobile devices, stacked on metal frames, created a technological assemblage of seemingly endless screen spaces. The phenomenon of group control systems was documented and introduced as a technological object. Videos and images appeared in a sterile and contextless form with a focus on the mobile devices, showcasing an autonomous system with no human intervention. The video, published in July 2019 on the short video platform Weishi, is one of the most re-shared videos of group control systems on Twitter. The recording shows an industrial hall in Shandong province of over 1,000 square meters with group control stations spread across the entire facility. With roughly 15 smartphones in a row, 4 stacked on top of each other, one station encompasses up to 60 devices, resulting in a total number of about 8000 phones, consisting of a 18x8 grid. But next to the stations the hall is empty, with half of the stations turned off, while the other screens light up as scattered pixels. It is important to note that the screens do appear to be static, showcasing non-moving images, a clear distinction to other later appearing visual material of group control systems. The inconsistency in the video is a chair which comes briefly into sight at the beginning of the video. A chair which symbolizes the hidden labor within the imagery of these systems and reveals the objects not as autonomous automated arrangements but as labor-intensive workplaces. autonomous automated arrangements but as labor-intensive workplaces. Although this work focuses on the emergence of a specific media arrangement as its research object and the phenomenon is certainly about its technological essence, from custom software to mobile device types and from SIM cards to LAN routers. Understanding media based merely on its technological capabilities and limitations lead to the fallacy of technological determinism. This is why this work finds great inspiration within the field of digital anthropology. This is why this work finds great inspiration within the field of digital anthropology. Digital anthropology is a branch of anthropology which demonstrates that the understanding of new digital worlds can only happen in the context of social relations and practices. Rather than being merely an instrument of finding arguments whether digital technologies have positive or negative implications, anthropology maintains a holistic methodology. Emancipatory Human Computation This work takes a cultural approach. It treats the lived experiences of people involved in the production of automated engagement and group control systems as the subject for theorizing the complicated structures of contemporary digital capital. Seeing how their cultural production on digital platforms and apps informed these people's labor and vice versa urges the understanding of such a technological object to a broader social cultural basis, a technological object to a broader social-cultural basis, looking beyond mere political-economic forces. Giving the scarcity of documented material, scientific writings and the hidden nature of group control systems, the approach of consequential and intimate research practices became an essential part of this endeavor. Human-driven condition combines a variety of methods, including participant observation within digital subpopulations, interviews with workers from different areas of the ecosystem, composing and contextualization of original visual material, and the positioning of the object within the functional of global digital capital. Besides the importance of these discrete methods, the essential part of getting insight to this vast technical assemblage was that I immersed myself within. With the reconstruction of an original group control system, I gained access to the spaces and platforms of the workers, allowing me to observe and analyze specific informal information, which were located at and were the essence of the system. This part of the research resulted in the contact with around 25 QQ and WeChat accounts in the form of unstructured interviews conducted via text, voice messages and video calls. It is by far the most valuable part of this endeavor but is by no means holistic or exhaustive given the time frame and the early stage of the research. Nevertheless, the indirect communication resulted in a way more meaningful result than expected. It resulted in hours of participant observation and gave access to further direct research. This part of the study was the involvement and observation of QQ chat groups in combination with official communication channels like short video platforms, Baidu, Taobao and WeChat business accounts. Given the circumstances of language barriers and no direct access to facilities, this inquiry is focused but not exclusive to visual material with a particular attention attention on workplaces. Let everyone have their own traffic. With the ever-increasing popularity of platforms like Taobao, Kuaishou or WeChat and their integrated paying methods, also the usage and ability to distribute products and services over them advanced accordingly. Now with well over a billion monthly active users, WeChat is a large part of the e-commerce system within China today. Through this demand, chat groups and live streams became the new place for transactions. Advertising services and products to multiple people at multiple places at the same time became a key element to the new way of distributing and consuming. The strong imagery of one person presenting products in front of roughly 50 smartphones staggered besides and on top of each other, all broadcasting to different groups and live streams on different platforms, while videos of anchor classes' rooms, filled with young people training specific presentation methods, all belonged to and are part of the business model that enabled the group control system. Because of this change in environment, a different iteration of this specific kind of model manifested, when next to the individual usage of multiple smartphones also a new business model emerged. This is where the currently most popular structure of the group control system materialized. In this iteration of the system, the smartphones moved from the desk to a parallel upright positioning of the worker's body. Workers could now operate up to 60 devices at the same time by stacking multiple devices in a row, with four arrays on top of each other. Within this iteration, the smartphones were not only used for processing tasks or managing local business operations, used for processing tasks or managing local business operations, but the operation of the sheer number of devices could be outsourced to independent suppliers. In these vast amounts of devices, the group control systems were often used to generate and sell traffic. Through the implementation of the group control system software, automated scripts executing on the phones were also introduced. Today the group control system is capable of controlling up to 120 devices from one computer. But the group control system itself is already an outdated form of its technology. Through the connection via USB, the scale of automation was limited by physical constraints. The current iteration of the object, which is called Cloud Control, makes use of WebSocket technology and connects the devices over a local Wi-Fi network with a single operating system. At this stage, the software changed from native applications to web applications, only sending instructions to a dedicated Android APK, which would execute the commands. This would allow the systems to add unlimited devices to the operation and would dissolve the physical coalition of devices and operator, resulting in imagery of computer rooms with devices numbers exceeding the multiple thousands. The largest assemblage of devices I encountered in my particular research was 20,000 by a company with its core philosopher being let everyone have their own graphic. Each studio has a different project. Each studio has a different project. Strikingly across all variants of businesses involving the group control system, the amount of custom and original software was astonishing. The variety of interfaces, scripts and methods was way higher than I anticipated. There was not one best way to do it or one monopoly which would produce the only available software. Every new studio I contacted that I shared original footage with their interface had similar structures and functionality to other software but was clearly an original product. Filling a niche within the market by offering specific automated tasks for particular apps ranging from the typical Douyin live stream interaction with custom words over the automated opening of red packets to the playing of mobile games or automated watching of advertisements. Because of the diverse use case of the object, Scriptwriter are one of the cornerstone to the vast ecosystem. Through the interface software study of Android Group Control System, one of the free versions of the operating software, I was introduced to the community of AutoJS, a programming library which makes the access to the Android API, easily accessible through writing JavaScript syntax. In these QQ groups, the netizens help each other with the automation of phones. They share programs, post tutorials, debug each other's code, and discuss the new layout of apps and workings of algorithms they want to access and circumvent. All the chat groups are in the size of up to 1,000 members. The groups have a constant activity, people replying to questions within seconds. When talking to Lee, someone from the group specializing on deceiving programmatic advertising, they explained to me how they program and release so-called carrier apps, which are essentially empty apps like simple calculators or random color generators to serve a platform for delivering advertisement. Lee would then download their own app and click on their own ads in order to receive money from the advertising platform. While explaining the technicality of circumventing the platform's anti-cheating mechanisms, they note, it's not just an automation system, more important is the business model. And they explain the vast hidden industry that sprung up because of group control systems. Each studio has a different project. Within automated engagement. Within the generation of automated engagement, the group control system shifts into a global context. automated engagement, the group control system shifts into a global context. Despite the importance of the specific locality and domestic usage of platforms within group control systems, it would be dismissive to not relocate it to the context of global digital capital, as these two spheres merge ever more. Chinese big tech such as Tencent, Alibaba, Baidu or ByteDance separates itself off from the Silicon Valley in distinct ways. But the firms are united in their efforts to produce technology that is oriented first and foremost towards the commodification of information. Within these extractive economic structures, the emergence of automated engagement serves as a particular subversion to the surveillance capitalist business model. One of the core functionality of large group control systems focusing on the traffic generation is the acclimatization phase of newly added devices. the acclimatization phase of newly added devices. This part of the automation can take up to one week and sets the foundation for any succeeding functionality. In order to not get shadow banned, the process is a vast complex simulation of random human behavior, including the scrolling through curated feeds, looking at profiles, following and commenting at times, but also the usage of other system-native applications or the importance of using the camera of the device itself to upload videos. In order to detect inauthentic activities, the platforms are using machine learning and access system-relevant information like phone contacts or text messages as home explains to me a work i talk to within the history of work automation has been the foundational mechanisms for dehumanizing workers and exploitative control of the working process. Here, however, automated alternative constellations materialized from the worker's struggle as an emancipatory bot culture. In digital capital's rationalistic workings, the worker is demanded to be human again, as the business model strives for authentic and genuine behavioral data. Maybe we can talk about that. Within the installation, the boundaries between research and artistic object blur. While the exhibited object contains the research object in its full functionality, the object also communicates its context and proposition, not only as an object but also as a method. Parts of the research entailing the worker's inquiry and gathering of descriptive content was automated within the group control system itself, creating an entangled media object as a system documenting and researching about, from and by itself thank you yeah applause I mean that's that's the thing that's now missing in these online conversations so if you could put some emojis in the chat that would be great yeah but thank you Conrad that was an amazing and amazingly short presentation I'm happy to talk more about it there are already some questions but we will keep the questions for after all the talks and then we will come back to Conrad to Conrad and the next speaker that I can announce is Marianne Lechner who will talk about and now the title is a bit difficult to pronounce XOXO DBFF star exclamation mark question mark emotional machine and Marianne is a specialist in user experience. She's a PhD candidate at the Kunstuniversität Linz and is also teaching here where I got to know her and currently she's doing research on IOT devices and the emotionality of these apparently smart living things. So I'm happy to give the voice to Marianne and I'm really happy to hear what she has to say. Thank you very much. So today I will tell you a little bit about my PhD project you already mentioned. And yeah, well, I can't get to the next slide. So since I can think of being in this world, I wanted to relate myself to the world. So as a child, I communicate with my family, with all the animals who were living in my home. I read the cherry tree next to my house poems. So mostly I was connected to them in an emotional way. And when in my childhood there was this series called Night Rider. And in this series, I think most of you will know this series, there was this special car. And the special car could talk to its owner. And they were friends, they helped each other. And I was so fascinated by these objects and I always wanted to connect with objects, talk with them, be friends with them. And this object could also communicate in an emotional way. So in my PhD I asked myself so what if our digital systems can show their emotion? The speculative design project uses a more than human approach to investigate the possibilities of inscribing the vibrant matter of emotions in the socio-technical interface in hcl so it investigates beyond typical concerns with products interactions and services and this discursive approach is to stimulate discussion and open up for debates. So please lean back, maybe close your eyes and imagine the following situation. So now it's 2040. You and your family are living in a flat in symbiosis with digital entities. Every entity plays a different role in your cohabitation. Your door is a friendly, courteous entity giving everybody a warm welcome and has a strong protective instinct. Your table is in the center of communication. It is smart and has a sense of humor but unfortunately tends to give private information out to its manufacturing company because it puts a lot of pressure on the table concerning its updates. So today is one of those days your door welcomes you with a warm color but the pattern in the left corner gives you already a hint that something is a bit different you enter your apartment and your attention immediately fall towards the table it seems like it has done something wrong did it upload your private data to the company again just to get an update for its interface it looks like a dog after stealing a piece of cake so in this scenario digital accents can show their emotion and react emotionally to a particular situation emotions are a part of the human experience and play a significant role in our lives, or maybe even more extensive than the rational one. Agency is a concept that has a long history in the field of interaction design and intelligence interfaces. Latour has tried to find a way out of the dichotomy between human-centered and object-centered notions of agency his term actance defines a source of action that can be human or not or not or not non-human or a combination of both something that acts or to which activities granted by others it implies no special motivation of individual human actors nor of human in general. Also, the theory of vibrant materialism from Bennett paints a positive ontology of vibrant matter, which dissipates onto theological dichotomies like life, matter, organic, inorganic and subject-object and sketches a political analysis that accounts for the contribution of non-human actions. What are emotions? The term emotion does not have a unified definition. This is quite hard to deal with. quite hard to deal with and especially in design there is no definition of emotion and according to the of the brand strategist Mariah Kibbutz who said that emotions are subjective emotional state that are primarily expressed through body language and serve interpersonal communication so this project is really focused on simulating emotions. We also distinguish feelings from emotions. So feelings is the inner state that what you can feel and emotion is the thing that you show to the outside. And most of the time you use it for matters of communication. So when I started my project, unfortunately there went something wrong with my picture by exporting the PDF. But when I started the project, I started by observing my coffee machine. So I was quite curious if I could already see some emotions so and i found some so sometimes it was tired sometimes sometimes sprout most of the time a little bit confused because it's an old one sometimes it was really angry because we didn't do the service for it and so on and so on and after this i tried to ask other people if they also find some emotion and i use the method of cultural probes so according to bill gaver cultural probes are an artistic designer approach it's openly subjective and only partly guided by any objective problem statement so it's a question or it's an input you can send to people and uh we use this data for inspirational data to stimulate our image imaginations rather than define a set of problems our imaginations rather than define a set of problems. So the cultural probe, How Are You My Digital Friend, was designed like a side of an emotional diary between things and humans. It introduces the participant to a vision of coexistence. People were asked, in the future we will live together with our digital devices, like in a family or community. Do you sometimes feel that your roommates show emotions or are in a certain mood? Can you observe any? And when? How do they think to show their emotions? participants between 10 and 25 years old, which objects of the private environment already exhibit special emotions and moods. Most of the participants draw an emotional case of an accident while others write from the perspective of the entity or as a reincarnation of themselves. The type of things encompass a broad range of heterogeneous things like a bad temperate shower, a fed up gold to cut, a mysterious but helpful cable clutter, angry mobile phones and PCs, lazy chairs, moody color pens, mirrors, and a jealous KitchenAid, just to mention a few. So I was asking myself what sort of emotions are relevant for interaction in this coexistence of human issues and what I did with this cultural prose, I tried to find out which emotions the things shows are more in the hostile way or in the friendlier way of this dominant emotions or submissive emotions. me one one thing that inspired me for the for the next step of my project so most of the entities we were living we we are living together they shown hostile and dominant emotions so in the next step I had to search for different emotions for a set of emotions and a repertoire of emotions. And I investigated a lot of emotional concepts. And the criteria of this investigation was on the third thing, the valence. So there must be a balance between positive and negative emotions and in the second step I also evaluated the potential for the HCI so here I found quite an interesting emotion models from me from the 90s so in this model it's really about the reaction. So emotions are not something just emerging. So there are some rational causes why we have emotions and most of. And we react on events, on actions, depending norms and rights. And we react on agents or persons itself. And from this emotion models, there are And from these emotion models there are different categories of emotions detected. Oh, I'm sorry, but I can't read. So we have emotions of empathy, of anticipation, of well-being, of attribution and attraction. So now I would like to show you a small project from Damme Reibi. It's from 2007 and this is a design fiction and also a speculative project. They wanted to investigate the new independencies and relationships which might emerge in relation to different levels of robot intelligence and capability. So they invented four robots. So one robot was very independent, another was very nervous, another one was sentimental or very needy. And what they found out is that depending on the robot or on the characteristic of this robot we tend to behave differently to those so they are shaping our behavior and they are shaping the way we we interact with them so one of the next things was okay i wanted to have a model of a relationship so there's an interpersonal relationship model from Kiesler and he also defines between hostile and being friendly or being a friend and a dominant relation partner and a submissive relation partner. And one example is, for example, when you have a very dominant partner, you tend to be submissive in order to have a good relationship or in order not to have a fight. So on the set, I put or located the different emotions from the set. And I think it's very important to say that emotions are something very cultural. So on the one hand it's a cultural thing and also an individual thing. So it's how do we experience emotion. So which emotions are allowed to show? So that differs from person to person, from family to family, from culture to culture. And this is something I will also expect in my interface that people can change the way the emotions are positioned. In order not to, so my goal is not to have a unified or global system of how this could be. The next step was, so according to this relationship I want to try it out, okay, so when I have a dominant hostile coffee machine what will it do when it has less water so what I detected this by defining different proto personas of this coffee machines it reacts differently depending on the repertoire of emotions it has so for example uh proto persona a which is dominant and hostile she gets angry or it gets angry and it hates me the prototype dominant and friendly probably gets a little bit aggressive the submissive hostile shows sorrow and fear and the submissive friendly one shows desire. So there are a lot of differences depending on the repertoire a thing could show. One other point of my PhD is how can we visualize those emotions. my PhD is how can we visualize those emotions. So one of my goals is to find a visual language to feel to find a visual translation and an abstract translation of those simulated emotions. And there's a structural similarity between emotions and formal language and for example many people draw anger anger the same way or a group of people draw engine in the same way so it's maybe hard uh it has some some uh it's it's not well formed it's not round or for example for enjoyment they use this this kind of signal so what I do for the next step is so I invite people to to show me their visualization of how would they draw stuff in order to find out criterias for defining parameters for the visualization of this emotion what i'm doing now is that i'm doing a generated application so when i'm with the cursor on an emotional landscape, I change parameters and try to find new forms of how this emotion can be visualized. So this is my working space now. So this is something I'm really doing at the moment. So I have not a finished project to show uh yeah but i'm really looking forward and i think the first cleanses so there are first things that you can see a difference between when something is really when something hates you versus something feels empathy or something feels desire for you feels empathy or something feels desire for you and well my question is how can digital systems communicate with humans in an emotional visual and related way in a more than human world applause again thank you very much Mariane that was also very interesting and very short and I'm looking forward to go deeper into emotional machines and technology maybe first of all is my microphone working alright again Can you hear me? Fine. Great. So lastly we will have a bot coming to join us or like a voice maybe. And this is an artwork from Kiriaki Guni which was exhibited at this year's Ars Electronica festival, and which was also part of a research project she did this year with Ars Electronica and others. And yeah, so we will watch a 27 minute lecture by a computer philosophizing about themselves. If the technical team can make it run. And then we will come back and the three of us will have like a little talk and answer and ask more questions than we could possibly answer probably together. you yeah another applause and um welcome back I think brings us into the exact right mood to talk a bit with each other even though one of the sentences that stuck most to me was you are humans and you are bound to get bored so I hope not everyone out there got bored and found it as fascinating and interesting to listen to this monologue from an algorithmic being talking about its own existence and talking to other algorithmic beings because as she said or as they said on the internet the biggest audience are algorithmic things that scrape and things that analyze our data that we put out there. Yeah so let's create more and this last call to arms to deceive and disturb these machine learning systems together. Maybe just a few quick words, Konrad or Marianne, do you have some thoughts on this last piece? Because I think it already has some... it ties a lot to both of your researches somehow. I think it already has some, it ties a lot to both of your researches somehow. Yes, it does, because I think it talks about the problematic of the difference on the one hand, and on the other hand, the difference of the separation of this human and objects, human and algorithms. So I think there is a need to overcome somehow. So I think there is a need to overcome somehow and there's also a need to find an authentic relationship or a private relationship or something different than in this capitalistic way how we use the things now so this is something and I also remember the sentence inventing new language because this is what my project is is also a little bit about so to invent a new language to communicate with the systems but also to to give them a voice to tell us something, a private voice to maybe tell us what they think about particular situations. Maybe I go already then deep into my critique of your research, because as you and all of us here are kind of seeing ourselves as designers, people who create objects and build objects for others to use, right? There's also this other discrepancy that we kind of make these tools and then as as you said also in your research that the emotion that people that people get when they use it is is very different from depending on the culture or how how they how they grew up and so how how would we maybe change this dichotomy where the designer and the user become one? How can we create objects that are kind of formable and that we can change depending on our own emotional state? I think we have to use parametric systems that are scalable. This is one thing and changeable by individuals. So this is also something I thought a lot because an emotion is something, if we hate somebody, I'm not sure if we hate the same way. So even so, but one time we said, okay, when we have to state something or when there's the situation, I show you hate. Or sometimes when I talk to my kids, yeah, I show them, okay, you have to do the homework. And I try to simulate a little bit of angriness or something. So I think probably it's also not necessary or maybe we shouldn't name those states or those reactions. Maybe it should also be a little bit more in an uncertain way. Or just to get an idea of something. So I think this is something I would like to go in this direction in my work. And if I understood you correctly, you're basically calling for more anthropomorphization. So like the coffee machine, because it is already somehow an object that has certain feelings, that tells you how to use it, that creates certain emotions, we should embrace this as designers, but also like culturally even more that each object surrounding us should get its own personality. And as maybe in the Kiriaki video would be interesting to think about, like should each object then also have its own voice? Like literally have its own voice that you can talk to everything. Is that the future where we're heading? And is that the future that is needed in your opinion? To be honest, I do not want to have objects with a voice because I think for me, this vision is a cacophony of different things who want to talk to me, so I think my future is not in the interaction voice because it's as she already mentioned it's too loud and I as she already mentioned it's too loud and I think what what so one thing why I do my my emotional thing is I would like people to have a little bit more respect on things so this human-centeredness of this world and also the human-centeredness in design. So I think we have to rethink this concept of humans being the only having the power of doing everything or things are only here to service humans. So when you look at the work, when you think about resources as the algorithm did, when you think about how the relationships are, how does the relationship... So we design, we shape objects and technology and algorithms, but also those objects, algorithms and things, they also shape us. So we have to be maybe a little bit more devoted or reflective what we do and what we design but also I think we have to yeah maybe maybe a little bit like this so that would mean that the coffee machine in your example would sometimes say like ah no I don't feel like I'm making you a coffee or like you have you had four coffees yeah it's too much yeah or probably say hey you didn't clean me yesterday no i i don't have any uh ambitious to serve you so maybe it's a little bit like this but what would this mean i was thinking a bit more globally that it like looks at the global production of coffee and how workers are you know for for making coffee beans are also exploited and then I was like ah no today like I don't really feel like making a coffee because of these political reasons yeah or it forces you to buy the right coffee yeah there's according like ties to this there's a question from the audience that is asking what is then the agency of objects with personality and I suppose this means like yeah with who who do these objects then serve if not the designers and not the user? I would like to ask the question, is everything about serving somebody? So is the whole world here to serve us? So this would be a question I would raise. I think you also mentioned Latour and its agency, but I don't remember quite that slide um so um agencies uh the accidents are just a source of action so they have this uh power to provoke some action and for us for example when when i have something like this it it shapes me that i do it like this to train from my team so we shape things but the things also shapes us and the things also are a source of action which influence us which which influence the situation for example for example when i when i put everything on this computer, you will not see me in a few seconds. And so everything is vibrant and everything can be a source of action. Yeah, there's another comment from Francis in the chat that says that this this new coffee machines like I mean we envision them maybe as like good good coffee machines but he has this doom vision that they will probably upsell you another product like in a Starbucks so you actually want an espresso but then it says like oh no maybe should get like a grande cappuccino latte um yeah but uh but i think i mean we talked a lot a lot about these coffee machines but we also have to talk about the invisible visible systems so i think that on the one hand there's this materialized object but there are a lot of invisible systems and Probably also have to think about so how can we rematerialize those systems just in order to get recognized by us Yeah, which brings me to another point of the materials of computing because in digital media we always try to hide that there's like some actual actual stones that we use to to uh to do computation on right and then i think that ties again then to conrad which i didn't that let's speak about his opinion on the um on the video of kiriaki goni um but maybe you can also go into these seemingly radioactive materials and maybe ask this very naive question of what kind of phones are they actually using. Did you find that out? Yeah, so Xiaomi most of the time because it's the cheapest and the most accessible also because they also build it all on used phones yeah that was kind of what I was expecting and then yeah and then they create like these massive systems of used phones that they connect to a central computing unit that then controls these phones or sometimes as you said there are like other actual humans doing labor touching scrolling yeah um and what what is the difference then between uh the term control group control systems and click farms is that the same thing? yeah I would say so group control system the name comes from a Chinese term I translated I also don't really know if it's a proper translation because most of the work is pretty lost in translation also, I guess. And I would say maybe click farms is a more broader term. Like the group control system is actually the actual object, like how we seen it with the phones on it. And the click farm can be also more distributed for example like how we also have it in germany that people you know like facebook pages for small amounts of money or like mechanical turk which are also which is also distributed click farm in a way of micro work and click work. And then group control system is kind of just a more specific term for this actual object, I would say. Yeah, and do you want to go a bit deeper into how this project started and how the first steps were to get in contact with people and which language you actually used to communicate with these workers yeah so um the the first contact was um how i described on twitter just these videos but then from from there used Google image reverse search. And then I was led to AliExpress and Baidu and Alibaba. So it was also a very commercialized way, let's say, to get to this project, because I actually got to shopping pages in the end, where you could buy the racks of the click farms. And from there, I... So then on AliExpress, for example, you have Wechat business accounts and here you can also make a wechat and like for outside of china and then this is how i got into contact with like more like official businesses that were running these like that were also selling these racks or selling used phones or selling just the USB hubs or selling just the cables or selling the routers. And then from there on, actually did um so i figured out okay wechat is more like this official way of contacting the people but qq would be way more personal because there are also these chat groups which are really like at the kind of like at the core of this running system and then i did a uh i had several contacts then on wechat and there i did a like a kind of cultural exchange of platforms where i made a facebook account for someone and they made they they approved my phone number on qQ with another Chinese phone number. And then I was able to, this went back and forth a couple of times, and then at some point I was able to use QQ and then from there I got into the group chats and then i saw what was going on there and from there i contacted people privately on qq and yeah because i just said to myself from the beginning okay i actually want to build this system myself i always said this entry point of like hey yo I'm actually not from China but I want to build one of these systems can you maybe help me somehow and I for the whole communication I use google translate and wechat translate and qq translate and to most of the people I was actually writing in German because they said yo just speak in your language i'm my app translates automatically so yeah so like what i'm always saying with like the work was lost in translation was also lost in translations of ai translating systems you know and yeah that's yeah that uh, so the language barrier was definitely there a lot of times, especially with the video calls I had with people, it was more, yeah, like, um, showing stuff and sharing my browser and trying to communicate like this so yeah and on the on the video calls were like did the people then know that you were doing kind of research on them or were they just showing you stuff how to rebuild these these projects did you reveal that you were actually anthropologically surveying them or yeah no i was i i thought so yeah for me it was important to just disclose this always in the beginning because then it it was also always feeling a little bit weird when i was talking to people which was their business of course and then after i don't know a week of riding they realized i don't even want to buy anything so in the beginning i always said yeah i'm researching it and i also actually want to build it myself and in the beginning i wanted to order things from china directly from them but then corona happened so this was not really an option anymore. And yeah, so most of the time they, like with the video chats, definitely they knew that I was just like this researcher, designer, artist who wants to understand their work, you know. And then actually at this point, it was always a very big shift. And then actually at this point, it was always a very big shift. Like when they understood that I was just like genuinely interested in their work in these systems, then the communication was always way easier than before me trying to explain something. And then, of course, a lot of times they were just like, what are you talking about? I don't understand. What do you want to buy? Just tell me what you want to buy because yeah the translation was a lot of times also just not working well i mean i mean this is super interesting where uh also i mean looking into uh again other other people sharing code online and using open source software and how how how open people are in sharing sharing their knowledge and kind of getting together and that it's like a cross a cross-boundary thing that like also bridges different cultures is very very nice to know and interesting to see and how far did you get in the end in building one of these systems or using, what was the AutoJS? How far did you get there? Yeah, so I managed to build a complete working ClickFarm with a specific software which was running as one of the more earlier group control systems so it was all locally and i had like a usb hub of 30 usb ports so i could run 30 phones and of course the platforms and the ecosystem around it was very different so because i i used i mean i downloaded do yin the apk so i was able to actually um not use tiktok but um of course the like i could not open now a shop saying hey i can actually sell you or offer you a product because it was not running yet as smooth as i wanted to so until now i used it more as a like how I exhibited it in the end like as my bachelor thesis it was like a more it showed its own context and was automating human behavior like it was just scrolling through Douyin or other apps and it was posting comments and stuff yeah so so this library is not even made for let's say tiktok or like western applications yeah so the chinese market yeah the no actually so auto js is just a way how to automate android phones and also in these groups there were you know a lot of different reasons why people were using this um they were also definitely um splitting communities like one group like specific groups were saying hey we are not in this business of group control systems we actually just want to automate for our small business our phones you know because just so much is running through phones in china and then there were groups which were really like way more focused on the financial side of it you know to to like uh yeah to generate money but i was really trying to not like i was really trying to not judge this in any way because i was generally interested in what the system is used for in in general and actually on the side of the more i mean it's all also a gray zone. And now the government started to ban businesses that are based on this, on merely this traffic generation. But at that time, when I researched, this side was way more interesting to me because it was way more about understanding the algorithms of these apps, of these platforms, and trying to deceive them in any way. Also, I mean, what the video now said, you know, deceiving and generating, like, of this whole thing we are stuck in, you know, to generate just so much data, to generate noise and to create a subversion through that. And I think it's still a very interesting approach. I'm just not sure. interesting approach I'm just yeah I'm just not sure I think this is also that like the point that struck me the most that there's this company that creates app an app that has advertisement in it and then they create the the bot farm to use the app to just get like all the all the revenue out of out of the ad system behind it right yeah exactly they create all these loopholes and i mean really i was just every day amazed by the amount of knowledge that goes into creating these systems if there's one more update of a game where you can farm cryptocurrencies or whatever it would take like only maybe one or two days and then there was a new script out and then people would buy that script from that person that originally made it and then they're really all these little jobs created in the system of like the people who write the scripts the people who more sell the the hardware or the people who are really just into not um not only just scripting but understanding the algorithms so they in these group chats they were always writing oh i noticed this behavior of the button of the like button in douyin is now two pixels more up or something and then everybody would update their scripts and stuff like this would happen like in intervals of some minutes so yeah and all human labor to understand this i mean you you focused now on this um Chinese or Asian specific uh side and why is that like did you did you find any groups uh working with click farms or automated systems in uh English-speaking countries or Europe or yeah so I mean this this focus now really just came naturally through the the imagery that I started to you know to research and of course this was also like a question I posed myself and until now I cannot really answer this until now I didn't find any of this these systems outside of China yeah like my ultimate thought also like while watching the last video that I mean was like this it was again like a very nice way of putting it so it's like this automated avatar that lip syncs whatever a person in way of putting it so it's like this automated avatar that lip-syncs whatever a person in front of a camera probably says and then and then there's a human person actually then delivering the words making it somehow emotional and then i don't know i was asking myself also seeing these these images where you had like an influencer with like 50 phones in front of them uh i don't know i saw this dim future of uh like these digital avatar influencers that already exist um but maybe also in relation to mariana's work maybe those are not then like 3d 3d avatars but they are like uh your fridge or your coffee machine who's trying to influence you or manipulating you and your human behavior. Lastly, I have a very hard question from Manuela that is asking about why do we mystify these processes and mystify machines and algorithms and why do we even develop them further to mystify them even more so i understand that question in in relation to um maybe also what i'm thinking to answer it myself maybe a bit is this um when i talk about my computer i try specifically not to anthropomorphize these objects and i try like the more you understand how how these machines work the less the more you understand that they are really different from from human beings and that also goes into account and even more so with like ai systems where we got into a territory where we use language that is hyperhuman uh that where we where we talk about things like this machine read thousands of articles or this machine read thousands of books and now can translate from chinese to english but that's that type of reading is completely different it's like a statistical analysis and i don't know from my side maybe I would say it's a call to action also for designers and producers of these kind of systems to Not mystify them as much but maybe that's not what the question is. How do you read that? Maybe Mariana? Why are we so much to need of mystification processes that we accept them and even develop them further? Why do we want algorithmic companions? I think it's really a hard question. And I think we... I mean it's like a hard question and i think we we um i mean it's it's like it it's like a tool so we are we as humans we are developing tools we we did it uh decades and tons of years and so when we invented the digital systems uh of course we we i mean we we get we we get to know a lot of information just in the glimpse of time so this helps us a lot it's it is servicing uh you can learn easier you you get uh informations you get everything so there's a big fat a big fascination of those machines i think we we always wanted to have a companion with different things so i think we we also always wanted to have this relatedness as one of the human basic needs to different things in the world. So this is maybe a second cause. And the third thing is the information the machines have. This is so overwhelming for us and also in your project so this uh so human becoming a a relay of an of a digital process so they are not not it's not that human like to work like this i would say like this so they're becoming kind of relate just to to do some uh to fulfill some actions and i think in the end we just know as humans and how we humans are seeing and experiencing the world so this is this is our starting point i think this is why we do it like this starting point i think this is why we do it like this conrad do you want to add yeah i i mean of course it's it's a really hard question and i would really really like to have an answer for this it's for me i still because i'm more coming from this programming background i mean i i directly think of how the open letter of bill gates trying to crush free software and open source to somehow put patents on code and put you know take ownership of systems and And I mean, of course, it's like capitalist realism, but for me, the only answer I have to this question is because small people want to be, like specific people want to be in control of these processes. And that's why they don't share the information you know that's why these systems are so black box and i mean ai of course i think is still like neural networks and deep learning i think is still maybe a different, would be a different approach to this because I think this is maybe not the result now because of commodification. But I think radical transparency transparency and I don't know if we are actually in need to mystify these more you know I think just if very few people are in need to do that the rest of us actually is not in need everyone is like why is everything so complicated while it's maybe not as complicated as it seems i'm not sure it's uh it's a tough one what do you think well well i think as you said i mean on the one hand we really need uh designers to make complex systems more usable and kind of trim it down. But then on the other hand, I would also love to see more work out there that teaches people like the inner states of machines and of complex systems. so rather than dumping it down to make it um i don't know in the 90s we had like manuals right they're like books uh that people would read in their home like why do we need like the reason why we don't need these manuals anymore is that the systems are so easy to use that we that they teach teach it by by touching different points and then um i don't know i think lasso would have added to this uh by like yes this when we talked the last time in the summer he had this term of like um um machine driven design or like um environments what we're creating is actually environments for machines. So these algorithms around us are actually teaching machines or teaching humans to act like machines. So to be more easily quantifiable and more easily readable. And I think also from you, Conrad, you had this sentence. I don't know if you took it from someone automation as ideology and i think this is something that i want to think about further and yeah i really like that automation as ideology um yeah but maybe if we we can close it down today I don't know what do you think I think it was a fruitful discussion and it kind of worked also online yeah it was nice it really actually felt like a conversation and if nobody has like any shouting online I don't know if there's still people watching. Maybe humans get bored and they check their emails while we talk. Yeah, so tomorrow at six we will have another conversations with computers. and other conversations with computers. Tomorrow we will talk about esoteric coding languages, which are like these fringe computer languages that are like far away from any commercial systems. They are deliberately used to make programming really hard or really strange or really funny. And we will have great guests if you scroll down you can see the list you can check the personal websites of all the all the participants from today but also from tomorrow and if we if you have more information if you have links just put them in the in the etherpad that you can see that you can maybe information, if you have links, just put them in the Etherpad that you can see, that you can maybe not see when you're using Chrome or a Chromium-based browser. But yeah, also Raki was saying it's very interesting that I'm very happy that she that she came and that she listened. I'm also really happy that we were able to show this video and it was super interesting to have this avatar as part of our kind of video conference which was really funny. Yeah, so far I think that's it from us. i will stay a bit in the chat and we can we can talk or i don't know go on check out the internet together or have a break maybe thank you so much mariana thank you conrad and uh thank you again antonio and thank you for being for setting up the streaming and um to fatima who was doing this amazing job of like putting all our putting all the notes together that we have now in a written form great boss Thank you.