DISCUSSION POSTS
FORUM 1:
wolchover and ai
​​
Wolchover describes in her article that faults occur in AI and robots when human instructions for reward functions are not clear enough, leading to a range of issues, from polarization in the Youtube algorithm to self-driving cars that are a little too cautious. Wolchover argues that these issues will resolve if, instead of ruthlessly serving a reward function, the robot seeks to determine human preferences. Of course, this has its own caveats, as we hardly know our own true preferences, preferences often change, and humans are not completely rational.
Stuart Russel, a Berkeley computer scientist, explained that robots should “simply try to improve our lives”. However, even if robots were to be able to successfully determine our preferences, robots are only dictated by those who have access to robots in the first place. And, I would argue that the preferences of those with the power and resources to control robots have preferences that frighteningly resemble a reward function, which is the exact scenario that Wolchover believed always led to a “misaligned AI”. Under our system of capitalism, people in power are essentially driven to make as much money as possible and keep costs low as possible, encouraging them to exploit people, the environment, etc. Corporate colossals such as Uber and Lyft are at the forefront of a tech-driven workforce and are a clear example of what happens when the capitalistic gain is paired with advanced AI. According to Robots were supposed to take our jobs. Instead, they’re making them worse, the gig economy is powered by AI, which “lets workers flood the app to compete with each other at a frantic pace for pay so low that how lucrative any given trip or job is can depend on the tip”. The AI also punishes workers who do not accept rides or who do not work at certain times of the day in order to serve the preferences of those in charge by bringing the most economic value for the people in charge. These gig companies even went as far as to severely inhibit worker rights through Proposition 22, something made possible with this new technology-driven, “boss-less” workforce.
Although robots certainly have the capability to improve our lives, with the social structures we have in place, it is much more likely that robots will determine the preferences of the top 1% than the general majority. Furthermore, Brandom’s Humanity and AI Will be Inseparable, a world is envisioned in which “it will be hard to distinguish human agency from automated assistance”. This world seems inevitable if the goal of robots is for them to determine our “true preferences”. The line between humans and robots is also blurred in Should robots have a face, as Walmart robots are even gendered, thrown birthday parties, and allowed to speak to customers. Maybe if simply trying to improve our lives solely consists of economic gain for a minority, then robots will be successful. But, that world seems more automated, homogeneous, and emotionless for me to consider it an improvement.
FORUM 2:
my research: tech and internet culture, gender and social norms
Newton’s Speak, Memory surprised me, as I did not expect to understand the point of view of someone who would preserve their friend through an AI chatbot--especially after watching that Black Mirror episode. Yet, Newton deeply characterizes Mazurenko as we learn he was intriguing, ambitious, unique, and had a charming way of bringing people together. When Newton writes about how sudden, how seemingly avoidable Mazurenko’s death was, we cannot help but feel empathetic and saddened. I understood why Masurenko’s friend, Kuyda, didn’t feel like she could compress the memory of her best friend into a “coffee-table book” or a “memorial website” and I understood why she would want to have one last conversation with him after he left so abruptly. Newton makes a strong case as to how digital footprints and advanced AI technology could help us grieve--serving as a listening platform for confessions.
However, the concern with any type of technology is: where do we draw the line? An AI chatbot may seem relatively harmless, but it could normalize people not only denying the death of a loved one, but also denying death in general because something so realistic can exist after they pass, or can even keep them “alive”. Klee’s Ever After is a bit less optimistic. Klee mentions a few dystopian novels regarding artificial intelligence and immortality, but The Possibility of an Island stood out to me. Klee describes the book as containing a “meaningless sort of immortality” as life stories are mindlessly passed on from clone to clone. To me, this use of AI feels like living in a literal sense, while not truly living. Immortality would take away the vague sense of urgency we have in living fully and experiencing it all because there will always be another day. I think that people seek immortality so they don’t have to experience loss, similar to what Kuyda did. But, if “death is what gives life shape”, then without the idea of loss, what is the idea of life?
I think that it’s hard for me to dictate how people should and shouldn’t grieve, and while I do think that Kuyda’s case was a bit unique since she happened to work for a similar AI startup and loss was felt so powerfully, it is impossible to stop other people from following suit and making more realistic replications of their dead friends and family--and eventually themselves. While the people who used Kuyda’s AI seemed to be respectful, Klee hints that immortality technology could easily be corrupted as the immortality movement seems to be led by people with large amounts of capital. Klee writes “should this ageless aristocrat come to power, we mortals would play the part of slaves and cattle, fit only to be sucked of essence, thereby extending his reign”. Even if advanced AI can help some people deal with loss, this could just be a baby step into something much more serious.
FORUM 3:
cryonics and ouija
In A Dying Young Woman's Hope in Cryonics, Kim Suozzi--a young woman with one of the deadliest types of cancer, tries to escape death in hopes that a new technology will one day bring her back to life. Kim’s story vaguely resembled Roman’s in Speak, Memory in which both were relatively young and had a lot of life to live before they died. In Roman’s story, his loved ones had a hard time dealing with the loss, so they recreated an AI bot with his text messages in order to converse with him. Kim, on the other hand, played a role in preserving her brain to be recreated in perhaps a much more intensified form.
Cryponics, AI chatbots of dead people, and Ouija are all similar in that people might turn to them in moments of hopelessness. In The Strange and Mysterious History of the Ouija Board, the author explains how speaking to spirits and the dead “offered solace in an era when the average lifespan was less than 50 (Links to an external site.): Women died in childbirth; children died of disease; and men died in war” and tended to peak in moments of uncertainty or hardship such as during World War II. I think it is similar with cryponics, in that when people seemed to be faced with death, they can turn to arguably extreme or rather supernatural measures. We see in Harmon’s article that Kim and her partner spent a lot of her last years thinking about and planning what would happen after she died instead of completely cherishing the time she had left because they were in such denial when faced with death.
Kim’s experience raised a couple of concerns and thoughts for me. Immedietly after she died, I also found it slightly disturbing how she was immediately given a procedure. Her friends and family essentially had to watch her die, and without time to grieve or say goodbye, she was rushed into a process that separated her body. I feel like if we are constantly seeking more time to live, we won’t actually process death, or even our last years of life. Kim’s procedure was also quite expensive, which raises a lot of equity questions about who gets a second chance at life, and how this extended life period could give relatively well-off people more power. Furthermore, similar to how people found ouija boards spooky for a period of time, I think it would be pretty natural for many people to find people brought back to life through cryonics, spooky. Would these people just be ostracized or would this whole process get normalized.