ZIYANG WU: INVESTIGATING THE ALGORITHM

Ziyang Wu: The most challenging part for me about working with artificial intelligence is constructing an interesting and conceptually challenging system to create a meaningful simulation. This ranges from using unique datasets to train AI, to juxtaposing a series of training models, such as those used in the video game industry, to reverse-engineering existing models that contain certain algorithmic or engineering biases.

I’m not interested in using any AI tool that comes with a pre-trained model. If you use a pre-trained AI, no matter how many interesting keywords you input or what kind of style it’s learning from, you are essentially just generating beautiful visual or audio experiences that are predetermined by others. The result is not going to be new or unique since nothing unprecedented has been created.

ZW: I don’t think the art world will allow "the end of human artists" to happen, as it will need this system to keep running. If we think of photography and the conversation around how it would bring about "the end of painting" in the past, for me, it is a similar discussion right now.

ZW: I think it depends on what kind of artist they are. If the artist uses AI as a tool like text-to-image generation, it could be used to generate ideas quickly, very similar to how artists use collages to come up with ideas. It is also a way to democratize art making, especially for those who are interested in making art but might not have the technical skills, like painting, to execute.

If the artist is a creative technologist, they will explore and invent new code, training models, and datasets, which could be shared as tools for more people to use. They will also further explore more interesting potentials of AI technology.

If the artist treats AI as a subject matter, they will focus on the problem that AI and algorithms generate. The artist will raise questions about it or propose alternative models to challenge or substitute existing ones. This is exactly what my NFT, A WOMAN WITH TECHNOLOGY, was trying to explore.

ZW: Whenever I’m using a so-called new technology, I’m trying to be very aware of its medium specificity, rather than only focusing on the spectacle it creates. As someone who is excited about (but also hyper-critical of) all technological developments, it not only gives us more tools to play with and more dimensions to explore but also the themes and topics to focus on.

ZW: The first time I minted an NFT was in May of 2021 when the NFT marketplace Foundation invited me to be a contributing artist. This was in conjunction with the launch of their 3D feature in which an artist could upload and mint a 3D model as a glb file and view it via Augmented Reality. Of course, I’ve been observing and following all kinds of news and events since 2019, but haven’t gotten involved until this launch.

NFTs offer so many new possibilities for artists to create and have a more sustainable practice. For example, let’s say there is a complex digital image that consists of 20 separate layers in Photoshop. We can mint each layer as an NFT, and authorize the collector with the right to change the layer they own. Every time a single layer is changed, the whole image changes. Since these changes are "real", encrypted, and irreversible, this "never-ending" simulation could reach its massive creative potential and offer a new sphere by merging art-making, collecting, and participation. Moreover, a lot of generative and live-simulation works have previously been difficult to sell. Many of these works live on the Internet and used to be only commissioned and collected in small numbers by bigger institutions. However, after being minted as NFTs, they become the "one and only" entities that are favorable to collect while keeping the uniqueness of their ever-changing nature.

I’m also very inspired by many NFT communities where artists, collectors, creative developers, curators, people from the tech world, etc. are regularly having community-based meetings and events both online and in IRL. I believe it will further bring people together and eliminate the hierarchies that often exist in the traditional art world.

ZW: By training an AI chatbot whose knowledge equals my knowledge, and making an animated video based on the content generated, "A Woman with the Technology" examines how an individual’s online experience becomes personalized by the Internet algorithm and creates the filter bubble. As a result, users become separated from information that disagrees with their viewpoints, effectively isolating them in their own cultural and ideological bubbles.

The project started with a 3-month long recording of all my online experiences. A series of keywords closely related to my daily focus is selected and used as initial input on various media platforms such as Instagram, Facebook, and YouTube. The keywords include censorship, surveillance, alienation, absurdity, Neocolonialism, post-truth, trade war, Asian experience, 5G, artificial intelligence, mixed reality, K-pop, and identity, among others. I log in to these media every day, giving likes and leaving comments on the posts and videos I’m interested in. The daily searching experiences are screen-recorded and organized.

At the end of the project, I used all the collected data to train an AI chatbot, who can be seen as an AI version of myself, or a self that was constructed by the Internet algorithm. The data that were used to train the AI include all the captions from each post, subtitles from YouTube videos, as well as all the textual information generated from the images and videos using another AI application called dense caption image detection. This part of the project was in collaboration with artist and creative technologist Yang Wang.

After the AI chatbot was trained, I googled "How to write a film script"? It listed a series of questions that needed to be asked to write "a successful script". I asked the AI chatbot all the listed questions. It then generated a futuristic and dystopian film script and I made a 12-episode animated video based on the AI-generated script.

ZW: I haven't lived in the United States for the past two years. Therefore, I don't think I have enough first-hand experience with election-related matters and may not be able to answer this question well.

However, false information has indeed become more unpredictable overall. Some time ago, I listened to a podcast that mainly discussed how AI-generated fake pictures, videos, and sounds are widely used in telecommunications fraud and the gambling industry. (As we all know, unfortunately, aside from the military sector, the early and most widespread applications of such technologies are basically in illegal fields.) In the past, we could employ biological information—such as fingerprints and facial features—the most "unique" information carried by each individual, for verification. But nowadays, even this information can no longer be truly "unique." As a result, engineers are once again developing detection tools that can identify these generated contents. Perhaps humans are like this: we continuously create "new" things, and then we need to expend more time and effort to create even "newer" things to counter what we previously created.

ZW: My artistic interests started with macro-alienation, which occurs on a global scale through political and economic oppression. My past multimedia works "The Story of the Pig" and "Carnival 2020" addressed this through slaughtered bodies and killed spirits, a world related to George Orwell’s "1984," and the critical examination of the Social Credit System – a "Black Mirror" reality. Other works refer to the invisible forces (such as algorithm-based commercials and data surveillance) that dissemble an individual’s value and rationality in their everyday lives or micro-alienation. In my animated video "The Last Subway," for example, I created the narrative of a fictitious video game protagonist ZiyangMon (Referring to Pokemon Go) who travels between mock newscasts, the Japanese pop music video PPAP, surveillance systems, and Virtual Reality footage to hint at how the virtual world increasingly influences human consciousness on a micro scale in our highly globalized and digitized society.

In the current post-Internet era, micro-alienation consistently appears with new faces and generates new dominating powers that have never been before. In today’s world, everyone is becoming a cyborg. Smartphones have truly become part of the human body. Manipulative algorithms are implemented everywhere: social media, search engines, streaming services, surveillance, the healthcare system, and so on. Alienation indeed reconstructs us more microscopically and is becoming more ubiquitous and harder to detect.

ZW: "Where Did Macy Go?" is an 11-episode animated film. The piece discusses the collapse and emergence of community structures after decollectivization, Confucian obedience vs. social obedience, the new tele-republic of home, “mask politics” and social justice under the pandemic. During the pandemic when everyone was simply exhibiting work online, I preferred to take advantage of what the Internet does the best: sharing and re-distributing. I posted all the videos on TikTok, one episode per week for a total of about three months. I was hoping to create this kind of excitement similar to the anticipation of a weekly TV episode. I asked all my friends to share it and even bought an advertisement and created Instagram filters and AR works, which were conceptual choices in favor of even further distribution. So far, the filters have been used over 300,000 times. In these choices, I am appropriating and “celebrating” an advertising strategy, while also critiquing it.

ZW: The initial inspiration for A WOMAN WITH THE TECHNOLOGY comes from the ubiquitous algorithmic manipulation in daily life. For example, right after you browse a product on Amazon, you will see similar product recommendations on Instagram the next second. Or the leftists in the United States rarely receive right-wing information on media platforms (and vice versa). Having lived in the U.S. for over seven years as an immigrant, I realized I was living and being "protected" under the "blue" and democratic "bubble" – one that I support and align with ideologically. I started to think about what kind of person I was re-constructed by this current bubble I’m in and how biased I could be without knowing people from other bubbles.

Another inspiration was from the book titled "Algorithms of Oppression" by Safiya Umoja Noble. The book's cover is a Google search bar with the text "Why black women are", and the suggested words are angry, loud, mean, lazy, annoying, insecure, etc. This is a very vivid example of how we are involved and affected by algorithmic bias in our daily lives. Then I started from an individual’s perspective, explored my current state of existence (or what the Internet algorithm thought about who I was or what was interested in), and reconstruct "my filter bubbles," to complete a deep research and exploration of the digital power structure.

The research and data-collecting phase of "A Woman with the Technology" were also very interesting. Beginning day two, I began to receive related posts and videos about Andrew Yang, the former Democratic presidential candidate for the 2020 U.S. election. Initially, I had no idea who he was and I think that the algorithm assumed I was Asian based on my browsing history. It thought I might be his potential voter, despite being ineligible, and would not vote for him. The idea and mechanism of "supporting the candidate based on the same race" are funny and ridiculous despite having functions.

ZW: If we are talking about the larger scope of Metaverse, such as the seven layers of Metaverse coined by Jon Radoff, countless aspects excite me. Or, to put it another way, we have been exploring all these aspects for a long time as artists, designers, and engineers. 5G/6G, micro-electromechanical systems, and the cloud are the Infrastructure Layer. Digital wearables and haptic gloves are the Human Interface Layer. AI agents, blockchains, and DAOs are the Decentralization Layer. VR, AR, and XR are the Spatial Computing Layer. Tools and assets for design and new workflows are the Creator Economy Layer. Social curating and rating are the Discovery Layer. Social media, shopping, and games are the Experience Layer.

ZW: On the surface, the phenomenon of algorithmic bias seems to have decreased significantly. This can be seen from the cases and reports of many large technology companies. On the other hand, this may also be because the phenomenon of algorithmic bias in the past was overly severe, prominent, and absurd, so it appears to be better now. But from another perspective, these "biases" have become more difficult to detect.

As for the phenomenon of algorithmic control, it must have become more profound (taking TikTok as an example), and it has become even more difficult to detect.

Our relationship with AI has changed from an initial state of confrontation and competition to one of symbiosis and coexistence because they (if we only consider them from the perspective of being tools) are getting better, more useful, and more ubiquitous. In this state, we increasingly "trust" the information they transmit to us and no longer feel discomfort and suspicion as we did initially. Then, those controls and biases that are increasingly difficult to detect may become even more dangerous.

Another huge concern remains monopolization. Monopolization here includes the monopoly of leading companies' businesses, the monopoly of infrastructure resources, and also the monopoly on the aesthetics of images in the future. Here, a comparison can be made; for example, there are only a few types of renderers for 3D software, which results in artworks made in 3D software being very similar in terms of aesthetics.

ZW: Yes, I have many "predictions," but this issue is just too vast. Even just focusing on the impact of AI on the creative field, I have many predictions. So here, I'll just focus on this part and share a few:

• The importance of AI agents will be extremely highlighted in any "creative field." We have already seen agents (through a variety of APIs) that can understand the creator's thoughts and carry out content output (including text, code, pictures, videos, 3D models, etc.) through multiple rounds of conversations. At the same time, these output contents are not isolated from each other but are interrelated and can even become input again to conduct content iteration once more. In other words, these agents are no longer just single-function closed-source tools like Midjourney or Runway ML. Instead, they are "creators" with a macroscopic perspective.

• Artificial intelligence will no longer be viewed merely as a tool, but as systems, prostheses, hallucinogens, collective unconsciousness, the Other, etc.

• Our relationship with it may evolve from denial to coexistence to symbiosis to battling.

• The end of truth (we are already deeply living in the post-truth era, and truth will only get closer to its end).

• The paradox of critique: There's a need to build entirely new ways to discuss concepts, ideas, texts, compositions, innovation, imagination, emotions, and so on. Here I'm envisioning an impending reality where all artworks could potentially be completed by AI or machines, ranging from photography/image creation to oil painting (via robotic arms), sculpture (3D printing), and of course digital artworks, and much more. Simultaneously, the concept could also stem from more sophisticated large language models (LLMs). If we find ourselves in such a reality, then most forms of critique will need to be "re-invented." For instance, if we see a remarkable image-based artwork today, we might say, "I like it because of its amazing color, composition, labor," etc. However, when we learn that it is created by AI, we might think, "Ah, AI did it and it's boring/annoying/not original," and so on. But soon enough, we might no longer be able to tell the difference. At that time, we might confront this new "paradox of critique," or we need to invent an entirely new way of "critique."

ZW: For me, the best NFT artworks so far, are the so-called crypto-native works that deploy the inherent features, rules, and spirit of blockchain. For example, all the generative NFT works that have been created using certain meaningful data, or that could be changed over time based on the collector’s interaction such as Pak’s "Merge" or Dom Hofmann’s "Loot." That is not to say these kinds of strategies didn't exist previously – digital and internet artists have been working this way for over three decades. But the scope is very different. A super successful crypto-native NFT project could attract many people and collectors to become a part of the work, while a socially engaged or social intervention net-art piece might be only looked at by a small group of people, and therefore the ultimate goal of the project might not be achieved. This is one thing that excites me the most about the next generation of NFT works: where all the infrastructures are better built, and there could be more socially engaged and meaningful work happening that could influence a larger group of audiences.

ZW: What excites me in the near future, is when more and better Web 3.0 infrastructure is built. Imagine when you purchase an NFT shoe, it could be imported in any Metaverse game, traded on any NFT marketplace, and “worn” via any augmented reality application. This is the future that Web 3.0 promises people to completely destroy the control from Web 2.0 giants. I very much look forward to this happening one day--although I’m not very confident.

ZW: Recently, I have been exploring AI generation applications such as Dreamfields 3D, where you generate textured 3D models from simple texts. Although it is still in an early phase, and I haven’t been able to train using a unique dataset, I’m very excited about its potential. Everyone could generate a complex 3D world/Metaverse simply using texts and from their imaginations.

ZW: After the initial hype of the first stage, when we revisit the seven-layer structure of the Metaverse today, the blockchain component, for instance, has lost some of its previous vibrancy due to various factors. Meanwhile, other elements such as GPUs, AI agents, VR/AR/XR, and games are experiencing rapid development. Generally speaking, I believe the Metaverse is an entity that is constantly evolving—advancing, retreating, and intertwining. In the future, we will undoubtedly witness even more changes.

In my work, I’m more interested in the whole (infra)structure and the complex connections and relationships between the layers. For example, Future_Forecast (in collaboration with artist Mark Ramos) is a project that includes a live simulation and collective world-building online environment, a series of NFT characters, and a CGI animated film (Pasig River 2030 – 6 Plus). In the context of “Digital Earth,” the “One Belt One Road” Initiative, and the Philippines’ “Build Build Build” Initiative, Future_Forecast presents and predicts the evolution of cloud-networked societies in the developing world. It does so by examining the impact of growing Internet-enabled networks and the ecological, geopolitical, and socio-cultural implications of a speculative future ISP and blockchain company in the Philippines.

In this project, we use Benjamin Bratton’s 6-layer Stack model (which is similar to the 7-layer model of the Metaverse) to focus on a developing region where the "superstack" platform has not yet been realized. By doing so, we attempt to simulate a future that bridges the virtual and the real, driven by both online and real-world data. By revealing the contradictions between digital structures and material reality, backwardness and progress, the project explores the impact of network infrastructure construction and development on developing countries in order to promote thinking about new infrastructure and discursive urban planning for sustainable development.