New day, new writing task. This is a reflection I wrote for the HCI course. I will tackle this in three parts, based on the references linked below:
The eyes are the windows of the soul, and apparently, your whole identity. This first paper discusses touches on a great topic: “how much does your gaze reviews about you?”. In the latest “Arctic AI Days” at Oulu University, there was a company sharing their research findings on early Alzheimer detection based on an ocular test. Only one lens in each eye, minimal data background, and someone will get to know the probability of having this disease. This information blew my mind, as I had close relatives that had suffered from the disease. However, I was truly impressed to find out that there is actually more information stored in our eyes. The fact that we can determine how much of a learning curve each individual holds and task prediction possibilities opens a field of improving current learning paths and making education more customizable.
As the paper went on, I was also fascinated by the amount of information that is stored in a person's gaze. I might not be alone in suggesting that optical scan identification is a new way to go in the future. This will further be explained in the semi academic section, but it is important to consider what we could use this data for, and how it can benefit the overall user. A possibility is to train algorithms to better detect sickness based on health patterns. A riskier possibility is to use this data to match two individuals. Who knows? It will be done for some interesting project research in the Applied Computing and HCI field.
Peru knows a few things about failed marketing campaigns. “Eager Sellers and Stony Buyers” was a pretty enjoyable read for me, given my former background in Business Administration. One of our nation’s prides is “Inka Cola”, this soda mix based on local herbs that stood up the Coca Cola company when they tried to invade the Peruvian market. As a result, Coca Cola ended up buying Lindley Corporation, proud owners of Inka Cola. In 2014, Inka Cola decided to launch a product that will “appeal” to Peruvians: a soda based on chicha morada, this purple corn and beloved drink of the whole nation. For the company, it was an excellent choice to diversify the portfolio and offer a new product range. For the consumer, it was a blasphemy to the “homemade” tag that has followed chicha morada since it is so linked to our heritage.
The opposite case is the one of group Añaños, creators of “Kola Real”, a soda brand that appeals to the economic sectors C, D and E of the population. This falls into a “smash hit” category since the brand managed to deliver a quality product at an accessible price. The product slogans say it all “calidad al precio justo”, which in English would be translated to “quality at a fair price”. I will close this section with a different product. Coca Cola company also has a market in Perú for juices (under “Frugos”), since our country is well known for its wide variety of fresh produce. In 2009, they had the great idea to introduce to the market “Hugo”, a new line that mixes fresh juice with milk. The slogan was crafted to introduce the juice idea in a cleverly written point. According to many marketing experts, the problem was that they tried to compete one-to-one against the freshly squeezed juice market. And as stated through this paper, this was actually inducing a significant behavioral change. In the eyes of the consumer, people will always choose freshly squeezed juice.
“Can AI without common sense be safe for humans?” The talk opens with a very interesting statement, and that has become a concern nowadays. Yes, LLM models have recently regulated their sycophancy levels after a wave of users turned to ChatGPT or Gemini asking for advice regarding personal matters. That has led to terrible outcomes like isolation and suicide. Since these models have been trained in raw data, in many times that means that the data itself is biased – has gender favoritism, racism and so on. I loved that the speaker made a call out that said data should actually be trained and judged by human tutors. There was another topic from my Social Computer course where we check this video channel with automated content that has childhood characters killing each other, and since the titles were SEO friendly, that is the first thing little kids will see popping up on their screen.
So... Common sense is not that common, apparently. And this is why most common-sense areas are ignored nowadays: there is more focus on logic, and we overlook the moral and normative areas to maximize revenue. Big manufacturers pushed their models over research laboratories that have no resources to train their own datasets. I certainly hope this issue gets more focus since there is plenty to explore still, as the speaker mentioned.
The second talk is an issue that I often see elderly people face: hear impairment. Through my volunteering years I met several people with disabilities including walking, vision impairment, and so on. Since I come from a third world country, getting access to operations or even equipment to normalize your condition is not common due to the cost of equipment. I would like to think that at one point technology will be accessible to everyone, so people like Rebecca can integrate into society and not face discrimination overall. As I watched the video, I also noticed most volunteering people feel this “patronizing” feeling not because of the person's condition, but because they perceived them as non-normal and tend to feel sorry about them instead of treating them as equals.
It also brought to my attention how communication channels change through generations. In this case, her example of “this generation hates voice messages, so they default to texting” was accurate, and I think it really benefits people with all kinds of impairments too. Text can be easily interpreted by these “alt-text” options and is friendly with people that are not able to speak and so on. The insight of “cochlear implant” was also novel to me, and she really debunked some questions I had about the implant function. Overall, this whole chat made me realize how many things are normal-user centric and not inclusive. I had spine surgery a while ago, and just then I noticed that most chairs are uncomfortable for someone that just recovered from surgery. At that point, I started wondering how many things are designed for practicality rather than usability.
The mapping through the jobs in the field of Ux was insightful. Indeed, there is no “one job fits all”. Amongst the ones that caught my attention were the “UX researchers” since they are in field people in charge of studying how a product is made and the “program manager” since I have experience managing a team. It also puts in perspective how many companies try to merge roles by having one or two people assuming 10 roles, and that translates into poorly designed products with limitations for other users. Following the second part of the video (the text), I like how each category has some research questions to improve their design mechanism, like “can this function cause motion sickness2 in the case of a VR designer and “what is the ideal language” for a conversational designer.
The second part of the UX course was informative as well. At one point, I thought the “universal approach” was a good design perspective since it covers most use cases. However, a “one size fits all” is a terrible perspective, and it is something that most fashion companies are adopting nowadays. Inclusive design seemed fair enough, with the “design for one, extended for many” focus that will cover most of the visual range users. But the equity-focus designed blew my mind. It is true: Some people need extra setup to use an application properly, so why not include it as a core part of the design?
Alas, it is important to have an equity-focused approach when we are handling datasets. As the video mentioned, when voice recognition appeared it was mainly recognizing male voices or English speaker ones, since that was the demographic it was trained on. That was poorly done and led to an extra delay in delivering a better service. I have designed plenty of web applications in the past, and the most common question is: “does this question need to be here?”. In most cases, people paste templates without considering why they were designed like that. Again, common sense is not that common.