My Summer as a Scyborg

Me and one of my co-interns in our last week at AWS.

In A Third University is Possible, K. Wayne Yang writes, "The agency of the scyborg is precisely that it is a reorganizer of institutional machinery; it subverts machinery against the master code of its makers; it rewires machinery to its own intentions." A scyborg, in other words, retools the technology of oppressive institutions against those same institutions for liberatory ends.

I arrived at Amazon's downtown Seattle campus for my software development engineering internship in Summer 2024 as a scyborg. I was pretty sure that I wasn't going to like my work and that I wouldn't be returning the following summer, but I came in ready to learn as much as I could about Amazon's internal machinery so that I could more effectively advocate for technology ethics after my internship.

To begin with, Amazon had not even asked me for team placement preferences, so I was already a little overwhelmed when I was assigned to the AWS Backup team that I knew very little about. I was even more overwhelmed when my mentor informed me in my first week that my project would involve modifying a codebase in C++, a language that I didn't even know. To put it briefly, my project was to create an exploratory proof-of-concept for a new storage mechanism to back up Elastic Block Store (EBS) chunks to S3 (Simple Storage Service) and measure the expected performance gains from doing so. I resolved to treat this internship as a 12-week long class and do my very best to learn as much as I could.

The first phase of my internship was to write a design document proposing a defending an approach to implement my project goals. I spent my first few weeks reading documentation and watching relatively boring videos about the minutia of Elastic Block Store architecture to get an in-depth understanding of the codebases I would be working with. Then, I compared different possible approaches along tenets like latency, division of duties, and extensibility. In writing my design doc, I quickly found that my philosophical writing skills were useful, helping me effectively communicate and argue for one proposed approach over others. Ultimately, my design doc was well-received, and I moved onto the implementation phase of my project.

Unexpectedly, I found that the biggest challenges I faced in the implementation phase was not the coding itself, which was relatively straightforward; it was all the infrastructural tools needed to run and test my code. Since I hadn't taken classes that involved working with command-line tools or extensive use of version control, I was lost when it came to many of the tools that my team used and had to learn fast. Through several weeks of late days at the office and hard work, I finally managed to get my proof-of-concept running.

I also learned a broad set of "soft skills" during my internship that helped me collaborate more effectively with others. I learned when to turn to other members of my team for help, and even members from other teams. My mentor also guided me on how to write messages that could be tailored to elicit responses from different audiences. By doing so, I was able to learn more about the idiosyncracies of the server codebase, for example, that were blocking expected performance boosts that we expected to be seeing. In the end, I learned many such soft skills that will be broadly applicable to anywhere that I go next.

At this point, you might be asking, what of the "scyborgian" part of my internship? Throughout my internship, I was always critically interrogating the possible ethical implications of my work. In particular, this ended up taking form as an investigation into the environmental footprint of my team's products, a focus that was inspired by my URBDP 200 honors ad hoc project and Atlas of AI, a book I read as part of HONORS 221 D (The Science of Human Values). As part of this investigation, I read both internal and external sources about the sustainability of AWS, and talked to the AWS Sustainabiltiy team about their work. I learned, for example, that the AWS Sustainability team was having a difficult time even just accurately reporting customers' carbon usage because they didn't have enough insight into the interactions between different AWS products. I tried in to address this knowledge gap by writing a sustainability primer for my team and including a section about sustainability in my final intern presentation, but ultimately I left pessimistic about the possibility of creating internal change and believe that real change will need to come from regulatory pressure. I also learned, however, that companies are not monolithic, and that there are plenty of well-meaning and incredibly smart people working within companies who can be allies in creating change.

As for my plans for next summer? I turned down my Amazon return offer and will be interning at a non-profit called USAFacts.