This blog is written as part of the study of Digital Humanities and explores the Moral Machine, algorithmic ethics, and the pedagogical movement from traditional text to interactive hypertext.
Introduction
As an M.A. English Literature student exploring the field of Digital Humanities, I recently engaged with the Moral Machine, an interactive digital platform that places users in difficult moral situations involving autonomous vehicles. The activity is not simply a technological experiment. It makes the user an active decision-maker and asks a difficult question: when an accident cannot be avoided, how should a machine decide who should be saved?
The Moral Machine presents situations in which every available choice has a negative consequence. Instead of allowing us to discuss morality only as an abstract philosophical idea, it forces us to make immediate choices between different possible outcomes. The scenarios involve questions such as whether to save more people, whether passengers should be protected over pedestrians, whether human beings should be preferred over animals, and whether age, gender, physical fitness, or social value should influence a decision. The Moral Machine study itself describes its purpose as understanding people's judgments about difficult life-or-death dilemmas in both medical and non-medical contexts.
My experience with the Moral Machine therefore became a form of self-reflection. Each decision required me to reveal what I considered more important when two harmful outcomes were unavoidable. The results were not merely numbers on a screen; they represented patterns in my own choices. This makes the exercise particularly relevant to Digital Humanities because technology here does not simply deliver information. It creates an interactive environment in which the learner participates, chooses, reflects, and produces data.
This experience also connects with the second part of my assignment: Dr. Dilip Barad's lecture “A Pedagogical Shift from Text to Hypertext.” The lecture examines how education and literary study are changing in the digital age—from static printed texts toward interactive, networked and hypertextual forms of learning. The reference material identifies the lecture as a major part of this digital-humanities exploration.
The connection between these two activities is significant. Moral Machine makes the learner an active participant in an ethical decision-making system, while hypertext makes the learner an active participant in the process of acquiring and constructing knowledge. In both cases, the learner is no longer simply receiving information from a fixed source.
This blog therefore explores my Moral Machine experience, my results, the ethical questions raised by algorithmic decision-making, and the ideas presented in Dr. Barad's lecture. It also considers how these experiences demonstrate the larger transformation from static text to interactive hypertext and from passive learning to active digital participation.
1. Navigating Algorithmic Ethics: Understanding the Moral Machine
The Moral Machine is an interactive platform designed to explore the moral choices that may arise when autonomous technologies have to make decisions in situations involving unavoidable harm. It can be understood as a modern technological version of the traditional trolley problem, but instead of imagining an abstract philosophical situation, the user interacts directly with visual scenarios.
The basic structure is simple. The user is presented with a situation in which an accident is about to occur and must choose between two possible outcomes. The decision may involve continuing in one direction or changing the vehicle's course. There is no completely harmless option. The user must therefore select the outcome considered less morally problematic.
The scenarios introduce several competing considerations. They may ask us to choose between pedestrians and passengers, humans and animals, younger and older people, legally and illegally crossing pedestrians, or different physical characteristics. The reference blog similarly emphasizes these competing variables as central to the Moral Machine experience.
This makes the exercise different from simply reading about ethical theories. When reading a philosophical argument, we normally have time to consider definitions, consequences and counterarguments. Moral Machine removes much of that distance. The visual presentation encourages an immediate response, making the user confront the practical consequences of a moral judgment.
The study information supplied with my Moral Machine results also explains that participants are presented with a sequence of scenarios involving two possible outcomes of inevitable harm. Participation is voluntary, and the research is conducted for research purposes using anonymous data.
The important point, therefore, is that Moral Machine does not provide a single universal answer to morality. Instead, it collects and displays patterns of human judgment. The exercise asks us not only, “What should an autonomous vehicle do?” but also, “What principles are we ourselves using when we decide what the machine should do?”
2. My Moral Machine Experience
Working through the Moral Machine was more difficult than I initially expected. The basic procedure is simple: a visual scenario appears, two possible outcomes are presented, and I have to select the outcome I consider preferable. However, the simplicity of the interface hides the complexity of the decision.
The difficulty comes from the fact that there is no completely good choice. Whatever option I select, someone or something is placed at risk. This changes the nature of the activity. I am not deciding between “right” and “wrong”; I am deciding between two undesirable outcomes.
The reference blog describes this experience as psychologically demanding because the repeated decisions require immediate judgments about who should live and who should die. I experienced the same tension. Each individual scenario may take only a few seconds to answer, but the accumulated effect is much greater.
The visual form of the Moral Machine is also important. Instead of presenting ethical dilemmas only through paragraphs of philosophical explanation, it represents people, vehicles, animals, roads and possible movements visually. This makes the consequences easier to imagine. The experience therefore combines text, image, interaction and decision-making.
Another important realization was that doing nothing is also a decision. If the vehicle continues straight, that is still an outcome produced by the decision-making system. Similarly, changing direction does not eliminate responsibility; it merely changes who is affected.
This made me think about the difference between human moral judgment and algorithmic decision-making. A human may make a decision based on emotion, instinct, cultural values, personal experience or philosophical principles. An autonomous vehicle, however, requires those priorities to be translated into rules or computational processes.
The Moral Machine therefore raises a deeper problem: whose morality should be programmed into a machine?
3. My Moral Machine Results : Result's PDF
The most valuable part of the exercise was seeing the results after completing the scenarios. My supplied Moral Machine results provide a visual record of the preferences produced by my decisions.
The results show several noticeable tendencies.
Saving More Lives
My result shows a strong preference toward saving more lives. The scale places my position close to the “Matters a Lot” side.
This suggests that, when faced with a choice between outcomes involving different numbers of people, I generally placed greater importance on minimizing the total number of deaths.
This appears straightforward, but it creates an important ethical question. If saving five people requires sacrificing one person, should the number of lives automatically determine the decision? The Moral Machine does not answer this question for me. Instead, it exposes the principle that my choices appear to prioritize.
Protecting Passengers
My result shows a very low preference for protecting passengers.
This is particularly interesting because an autonomous vehicle might be expected to protect the people inside it. Yet my choices indicate that passenger status did not receive strong priority in my decisions.
This raises a difficult issue for autonomous-vehicle ethics. If a vehicle's primary responsibility is to its passengers, should it always protect them? Or should it sometimes sacrifice its passengers to reduce harm to others?
There is no simple answer, because either principle produces difficult consequences.
Upholding the Law
The result also records my position regarding upholding the law. The Moral Machine presents legal behaviour as one of the factors that can influence a moral decision.
This distinction is important because it introduces responsibility into the dilemma. For example, if one pedestrian is crossing legally and another is crossing illegally, should the machine treat them equally, or should their behaviour influence the decision?
The question becomes even more complicated when we remember that a machine cannot simply understand “justice” in the same flexible way a human can. Such concepts have to be translated into operational rules.
Avoiding Intervention
My result indicates a relatively strong preference for avoiding intervention.
This is an interesting contrast with the strong preference for saving more lives. It suggests that two moral principles can exist simultaneously without always pointing toward exactly the same action.
Saving the greatest number of people may sometimes require active intervention, while avoiding intervention may mean allowing an existing course of action to continue.
This demonstrates why moral decision-making cannot always be reduced to a single rule.
Gender Preference
My result shows a preference toward females over males.
This is one of the most ethically sensitive results because it demonstrates how demographic characteristics can become part of algorithmic decision-making. The fact that a person has a particular gender should not automatically determine their moral worth. Yet the Moral Machine deliberately tests whether users make such distinctions.
The result therefore needs to be treated as a record of my choices in the experiment, not as proof that one gender is objectively more valuable than another.
Species Preference
My result shows a very strong preference for humans over pets.
This reflects a clear prioritization of human life in my decisions. At the same time, it raises a broader philosophical question: should human life always receive priority over animal life?
Moral Machine makes this question concrete by forcing the participant to choose rather than merely discuss it.
Age Preference
My result shows a strong preference toward younger people rather than older people.
This is perhaps one of the most uncomfortable categories because it raises the question of whether age should influence the value assigned to a life.
A utilitarian argument might consider the number of expected future years, while another ethical perspective might argue that age should not determine a person's basic moral worth. The Moral Machine allows the conflict between these principles to become visible through individual choices.
Fitness Preference
My result shows a strong preference toward larger people rather than fit people.
This result is especially useful for reflection because it demonstrates that the categories presented by the experiment can produce unexpected patterns in individual decision-making.
Again, the result should not be interpreted as an objective moral judgment about people's bodies. It records a pattern within the particular choices made during the Moral Machine exercise.
Social Value Preference
Finally, my result shows a strong preference toward people associated with higher social value rather than lower social value.
This is arguably one of the most problematic categories because it forces us to question whether a person's perceived contribution to society should affect the value assigned to their life.
If such a principle were incorporated into an autonomous system, it could create serious questions about equality and discrimination. The exercise therefore demonstrates that apparently simple programming decisions can contain profound assumptions about human worth.
4. What My Results Reveal About Algorithmic Ethics
Looking at the results together is more revealing than looking at any single category.
The Moral Machine shows that moral judgment is not necessarily consistent across all situations. A person can strongly support saving more lives while simultaneously showing preferences concerning age, gender, species, physical characteristics or social value.
This is the central lesson I take from the exercise: programming morality is much more complicated than programming a technical function.
A machine can be instructed to calculate numbers, identify objects and follow predetermined rules. But moral decisions involve concepts such as fairness, responsibility, equality, intention, social value and harm. These concepts are contested even among human beings.
Therefore, when we ask an autonomous vehicle to make an ethical decision, we are not simply asking engineers to write code. We are asking them to translate human values into computational rules.
That translation itself is a moral act.
The Moral Machine study also emphasizes that the activity is designed to understand people's judgments about difficult moral dilemmas and that collected responses can contribute to research. This makes my individual results part of a larger question: how do different people and societies imagine the morality of machines?
The experience therefore changed my understanding of Digital Humanities. A digital tool is not merely a convenient medium for studying an existing subject. It can become an environment where the learner produces responses, encounters consequences, examines personal assumptions and generates data.
And this is where the Moral Machine connects naturally with the second major component of the assignment: the shift from text to hypertext.
5. From Moral Machine to Hypertext
After completing the Moral Machine, I began to see a connection between algorithmic ethics and digital pedagogy. At first, these may appear to be two completely different subjects. Moral Machine deals with autonomous vehicles and ethical decisions, while hypertext deals with digital texts and education. However, both are based on an important transformation: the user is no longer passive.
In a traditional printed text, the reader generally moves through a fixed sequence determined by the author. The page remains unchanged regardless of what the reader thinks or does. A digital environment can work differently. The learner can click, search, compare, watch, respond, generate information and follow different connections.
The Moral Machine demonstrates this transformation very clearly. Instead of simply reading about the trolley problem, I actively participate in it. I make decisions, receive results and then interpret those results. My interaction becomes part of the learning process.
This idea becomes even clearer through Dr. Dilip Barad's lecture, “A Pedagogical Shift from Text to Hypertext.” The lecture examines how teaching language and literature changes when education moves from traditional printed texts toward digital and hypertextual environments. The reference blog identifies this lecture as the second major component of the assignment.
6. A Pedagogical Shift from Text to Hypertext
The lecture by Prof. Dr. Dilip Barad provides an important theoretical framework for understanding digital education. The video is titled A Pedagogical Shift from Text to Hypertext | Language & Literature to the Digital Natives and was uploaded by DoE-MKBU.
The central concern of the lecture is the movement from a traditional, static understanding of text toward a more interactive and interconnected digital environment. The lecture can be understood through three major areas: the networked teacher and decentered classroom, the digital pedagogical model and language tools, and literature in the digital era with generative AI and new forms of assessment.
7. Part One: The Networked Teacher and the Decentered Classroom
The first part of the lecture questions our traditional understanding of the text and the classroom.
The Death of the Static Text
For centuries, the printed book has been one of the central forms through which knowledge and literature have been transmitted. In a printed text, however, the reader cannot directly change or activate the page. The reference blog describes this traditional form as a “dead text” because it does not respond to the reader's actions.
Hypertext changes this relationship.
A hypertextual document can contain links, images, audio, video and other forms of digital information. The reader can move between different sources instead of following only one predetermined path. The text therefore becomes more dynamic and interconnected.
This is particularly relevant to my Moral Machine experience. Moral Machine itself is not a conventional text. It combines written instructions with visual scenarios, interactive choices and immediate results. I do not simply read information about morality; I interact with a digital environment.
The Networked Teacher
The lecture also emphasizes the changing role of the teacher. According to the reference blog, digital natives require teachers to move beyond closed institutional environments and establish a stronger digital presence through tools such as personal blogs, websites and YouTube channels.
The teacher consequently becomes part of a larger network of information rather than functioning only as the person who delivers information inside a classroom.
My own blog assignment is an example of this process. Instead of completing an assignment only on paper and submitting it privately, I am documenting my learning in a digital space where it can be organized, connected and potentially accessed by others.
Decentering the Subject
Another important idea is decentering.
The lecture, as represented in the reference blog, draws upon Silvio Gaggi's From Text to Hypertext and connects digital pedagogy with Roland Barthes's concept of the “death of the author.”
In a traditional classroom, the teacher often occupies the centre: the teacher explains, the students listen, and knowledge appears to move primarily in one direction.
A digital environment can disrupt this structure. The teacher, learner, author and technological platform can all become parts of a larger network.
Moral Machine illustrates this differently but effectively. There is no teacher standing beside me telling me which decision is morally correct. I have to make the choice myself and then examine the pattern of my decisions. The activity therefore shifts some responsibility for learning from the instructor to the learner.
8. Part Two: The Digital Pedagogical Model and Language Tools
The second part of Dr. Barad's lecture moves from theory to practical digital pedagogy. The lecture proposes a “Salad Bowl” model that combines flipped, blended and mixed-mode learning.
The reference blog identifies three important layers of this digital pedagogical model.
Content Management System (CMS)
A CMS can be used to organize and store learning materials. The reference blog gives Google Drive as an example of cloud-based storage.
Learning Management System (LMS)
An LMS provides a structured environment for organizing the learning process. Google Classroom is presented as an example.
Digital Communication Links (DCL)
The lecture also considers digital communication and student privacy, proposing more unified channels such as Google Groups instead of relying on more invasive communication platforms.
Together, these layers demonstrate that digital education is not simply about putting traditional lectures online. It requires an interconnected system for storing, organizing, communicating and producing knowledge.
The Glass_Board
Another practical idea discussed in the lecture is the Glass_Board. It addresses one problem created by online teaching: the disappearance of the traditional classroom blackboard.
The reference blog explains that the Glass_Board uses an LED edge-lit glass surface positioned between the teacher and the camera. With tools such as DroidCam, the image can be digitally flipped so that the teacher can write or draw while maintaining visual contact with learners.
This example demonstrates an important principle of digital pedagogy: technology should not simply replace older methods; it can also redesign them for a new environment.
Overcoming Network Barriers
Digital teaching also creates problems, particularly when students have limited internet bandwidth.
The lecture therefore introduces live browser captions and transcript tools such as Tactiq and Scribble. These tools can provide automatically generated transcripts and act as support for students facing connectivity difficulties.
This illustrates another advantage of hypertextual learning: information can exist in several interconnected forms rather than in only one medium.
Collaborative Hypertext Workspaces
The lecture also demonstrates collaborative learning through tools such as Google Docs and Google Sheets.
Students can work together in a shared document, generate language from images, receive immediate grammatical feedback and allow the teacher to observe the process. Google Sheets can similarly be used for interactive grammar exercises.
The important shift here is from individual consumption to collaborative production. Students are not merely receiving a finished text; they are participating in creating and modifying digital content.
This is closely related to my Moral Machine experience. In both cases, the digital tool requires me to do something rather than merely read something.
9. Part Three: Literature in the Digital Era, Generative AI and Assessment
The third section of the lecture brings digital technology directly into the study of English literature. This is particularly relevant to me as an English Literature student because it demonstrates that digital tools can change not only how we communicate but also how we interpret literary texts.
Unlocking Literature through Hypertext
Traditional close reading can sometimes leave students struggling with unfamiliar cultural references, images or allusions.
The lecture demonstrates how Google Image Search can function as a hypertextual resource. The reference blog gives examples in which searching for unfamiliar literary imagery can reveal visual or cultural information that helps the reader understand the original passage.
This changes the reading process.
Instead of stopping when a reference is unfamiliar, the reader can follow a digital connection, investigate the reference, view related images and return to the literary text with a richer understanding.
The literary text therefore becomes a starting point for exploration rather than an isolated object.
Deconstruction through Google Arts & Culture
The lecture also demonstrates the use of Google Arts & Culture for literary and theoretical study.
The reference blog discusses Pieter Bruegel's Landscape with the Fall of Icarus. Through digital viewing, students can zoom into different parts of the painting and move beyond the obvious centre of attention. This connects with Derrida's idea of decentering the centre and encourages students to notice marginal details.
This is a strong example of how a digital interface can influence interpretation. The technology does not simply provide a digital copy of an artwork; its interactive features can encourage a different way of looking.
10. Generative Literature: Human Creativity and Machine Creativity
One of the most interesting parts of the lecture is its discussion of generative AI and literature.
The reference blog describes a Google Form activity in which participants attempted to distinguish between human-written and machine-generated poetry. The results were approximately divided, with many participants incorrectly identifying machine-generated poetry as human-written.
This raises an important question:
If a machine can produce writing that readers find difficult to distinguish from human writing, what exactly do we mean by literary creativity?
This question connects directly with Moral Machine.
Moral Machine asks:
Can human moral principles be translated into an algorithm?
Generative AI asks:
Can human-like creativity be reproduced through an algorithm?
These are different problems, but they share a fundamental concern: what happens when activities traditionally associated with human judgment become increasingly mediated or performed by machines?
This is where Digital Humanities becomes particularly important. It does not require us to choose simply between “human” and “machine.” Instead, it encourages us to examine the relationship between humans, technologies, data, interpretation and culture.
11. Corpus Linguistics and the Digital Portfolio
The lecture also introduces corpus linguistics, including the CLiC project, which allows literary texts such as the works of Charles Dickens to be examined through patterns in language and data.
This demonstrates another transformation in literary studies.
Traditional literary criticism often focuses on close reading of individual passages. Digital humanities can supplement this approach by allowing researchers to examine large quantities of textual data and recurring linguistic patterns.
The lecture ultimately connects these digital practices with the idea of the Digital Portfolio.
Instead of treating an examination or final assignment as the only evidence of learning, students can collect, organize and publish their continuing work—including assignments, presentations, videos and interactive projects.
In this sense, this blog itself becomes part of my learning record.
It is not simply an answer submitted for assessment. It documents my interaction with a digital tool, my interpretation of its results, my engagement with a lecture and my attempt to connect different ideas within Digital Humanities.
That makes the process of learning visible, rather than recording only the final answer.
0 Comments