Qualitative Research Methods and Student Experiences of Artificial Intelligence

Research Context and Assessment Architecture

• Academic Framework and Course Architecture:

  • The study materials originate from module PSY1013 (Research Methods Stage 1) within the School of Psychology at Newcastle University.
  • The empirical inquiry centers on the central qualitative research question: "What is the student experience of AI?" (alternatively framed as "What are students' experiences of AI?").
  • Methodological approach: The research employs an inductive, semantic, critical realist thematic analysis applied to semi-structured interview data gathered from undergraduate psychology students.

• Summative Assessment Components and Strict Word Limits:

  • The final assignment comprises a comprehensive qualitative lab report accompanied by an integrated reflective log, submitted as a single unified document to Canvas.
  • Reflective Log: Capped strictly at $250\,\text{words}$. It mandates self-reflection on the principles of open science and an honest evaluation of how the student's perspectives on psychology have evolved—or remained static—over their first $6\,\text{weeks}$ of university studies.
  • Data Analysis Section: Capped strictly at $250\,\text{words}$. It requires a precise, systematic operational account of the analytical procedures undertaken, documenting the application of Braun and Clarke's reflexive thematic analysis across familiarisation, initial coding, candidate theme generation, and systematic code/theme review.
  • Results Section: Capped strictly at $750\,\text{words}$. It must introduce the generated themes, provide a clear thematic map, and present an illustrative write-up of $2$ to $3$ rich, conceptually deep themes.
  • Combined Word Count Scope: The combined total for the Data Analysis and Results sections is $1000\,\text{words}$. Direct participant quotes embedded within the results narrative count toward the word total; however, figure titles, table headings, figure notes, and thematic diagrams are excluded from the word tally.
  • Formative guidance is provided iteratively by teaching demonstrators within scheduled laboratory workshops, with reading week scheduled to finalize drafts.

Philosophical Underpinnings of Qualitative Inquiry

• Epistemological and Ontological Taxonomy:
  • Epistemology: The philosophical study of knowledge—interrogating its nature, possibilities, grounds, scope, and the validity of truth claims regarding how knowledge is created or acquired.
  • Ontology: The philosophical study of the nature of being, reality, and existence—examining what entities exist and the structure of reality itself.
  • Critical Realism: The foundational philosophical synthesis adopted for this analysis. It pairs a realist ontology (asserting that an external, objective world exists independent of human cognition or perception) with a constructivist, relativist epistemology (recognizing that all human knowledge of that external world is inevitably socially, culturally, historically, and conceptually mediated).
  • Reflexivity: The sustained, critical self-scrutiny by the researcher of their personal situatedness, theoretical preconceptions, potential biases, and assumptions, actively tracing how the researcher's presence and values influence data generation, interpretation, and analysis.
  • Thematic Analysis: A foundational qualitative research method designed for systematically identifying, organizing, describing, analyzing, and interpreting overarching patterns of shared meaning (themes) across an empirical dataset.
  • Inductive vs. Deductive Logic:
  • Inductive Approach: A rigorous, bottom-up, data-driven methodology wherein codes and candidate themes are formulated directly from the empirical text without attempting to fit data into a pre-existing coding frame, preconceived theoretical schema, or prior published literature. This approach is prioritized when exploring emergent technologies to capture novel, unexpected phenomena.
  • Deductive Approach: A top-down, theory-driven methodology wherein data coding and interpretation are guided primarily by existing conceptual structures, theoretical models, or predefined hypotheses.
  • Semantic vs. Latent Meaning:
  • Semantic Focus: Coding targets explicit, overt, surface-level meanings articulated directly and verbatim by participants, prioritizing their conscious representations.
  • Latent Focus: Coding digs beneath overt statements to interrogate underlying assumptions, theoretical frameworks, systemic ideologies, and structural conceptualizations shaping the discourse.

Reflexive Thematic Analysis Protocol

• Stage 1: Immersion and Data Familiarisation:

  • Researchers engage in deep, repeated reading of interview transcripts and active listening to audio recordings to comprehend participants' experiential lifeworlds.
  • Analytical prompts utilized during familiarisation:
  • Sense-making mechanisms: How does the participant construct meaning around artificial intelligence?
  • Etiology of interpretation: Why does the participant conceptualize artificial intelligence in this particular manner?
  • Interpretive plurality: What alternative or competing ways does the participant utilize to conceptualize the phenomenon?
  • Normative alignment: To what degree is the narrative conventional, common-sense, or socially expected?
  • Researcher positioning: How would the researcher experience the identical circumstance, and what personal biases or perspectives explain differences in emotional or practical reaction?
  • Ontological presuppositions: What fundamental assumptions regarding society and technology are implicit in the participant's worldview?
  • Revealed lifeworld: What kind of socio-technical reality is projected through the participant's narrative?
  • Core exploratory focus: Exactly how are university students deploying AI in their academic trajectories, and what discrete variables dictate whether a student experiences AI as advantageous or adverse?

• Stage 2: Systematic Data Coding and Review:

  • Codes serve as concise descriptive or interpretive labels applied to discrete chunks of data that directly answer the research question. They provide sufficient contextual granularity to permit the identification of recurring patterns across the dataset, and can be shorter or longer than the raw quotation.
  • Comparative Code Formulation Analysis: Evaluated against Participant 1 from Transcript 1 ("Erm, so it was kind of more an anxiety and a fear of the uncertainty and where our place is as humanity and that uncertainty regarding the future of it"):
  • Code 1 ("Emotions around AI development"): Deficient due to excessive generality; fails to identify the specific emotional valence or substantive core.
  • Code 2 ("Anxiety and fear around the uncertainty of the future of humanity due to the developments in AI"): Highly effective; accurately captures both the exact affective response (anxiety and fear) and the existential driver (humanity's future uncertainty).
  • Code 3 ("Anxiety and fear around the future impacts of AI on humanity"): Acceptable and succinct, yet captures slightly less nuance regarding radical uncertainty.
  • Operational Mechanics in Spreadsheets: Open a Microsoft Excel spreadsheet with dedicated primary headers labeled Code and Quote. Extract direct passages answering the research prompt and assign one or multiple codes.
  • Two-Tier Coding Review:
  • Part 1: Review codes across transcripts to verify grounding in multiple data points. Eliminate singleton redundancies, clarify ambiguities, and merge synonymous codes into unified labels. Preserve original audit trails by duplicating data sheets or saving distinct iterative versions.
  • Part 2: Return to full transcripts to recode the complete dataset systematically against the refined codebook, iteratively verifying the coherence between text excerpts and analytical labels.

• Stage 3: Theme Development and Visual Clustering:

  • A theme is not an inventory of topic responses, nor is it a summary of an interview question. A genuine theme must possess an overarching central organizing concept that unites diverse empirical observations around an underlying, shared idea.
  • Theme Verification Checklist Criteria:
  • Does it provide a meaningful, direct answer to the central research question? (Mandatory: Yes)
  • Does it merely summarize an interview topic or question? (Disallowed: No)
  • Does it encapsulate a broad spectrum of data united by an underlying core idea and supporting empirical evidence? (Mandatory: Yes)
  • Is it distinctly demarcated with clear boundaries separating it from companion themes? (Mandatory: Yes)
  • Does it maintain its own unique, robust central organizing concept? (Mandatory: Yes)
  • Is it constructed out of incoherent, randomly contradictory propositions without conceptual integrity? (Disallowed: No)
  • Clustering Methodologies:
  • Physical Paper Sorting (Method 1): Physically cutting printed codes into individual paper strips, spatially sorting and clustering them on a wall or desk, and labeling emergent clusters with colored sticky notes.

• Thematic Concept Mapping (Method 2): Developing visual mind maps that diagrammatically link central motifs to subordinate codes and empirical properties, visually tracking relationships and overlaps.

• Spreadsheet Color Coding (Method 3): Deploying the fill colour tool within Microsoft Excel (Home > Font) to assign distinct categorical colors to related code rows, maintaining an analytical color key.

• Stage 4: Candidate Theme Appraisal and Refinement:
  • Systematically interrogate candidate themes against the empirical dataset:
  • Do patterns capture authentic participant meaning without distortion?
  • Are themes internally coherent while remaining conceptually distinct from each other?
  • Are inclusion and exclusion boundaries clearly articulated?
  • Is there sufficient rich evidence across participants to justify each theme?
  • Does the theme reveal insightful interpretations rather than superficial descriptive frequency?
  • Warning against decontextualization: Progressive stages of thematic abstraction distance the researcher from primary data; regular cross-referencing against raw quotes prevents misrepresentation.
  • Formalize theme definitions by documenting:
  • Proposed Theme Name.
  • Central Organising Concept.
  • Conceptual Scope and Boundaries (explicit inclusion and exclusion criteria).
  • Distinctiveness (what is uniquely captured by this theme).
  • Substantive contribution toward resolving the overarching research question.
  • Concise executive synthesis summarized in $1$ to $2$ sentences.
  • Construct a formal Thematic Map: A diagram detailing all final themes, sub-themes, and intersecting structural relationships, produced via digital design tools (Canva, Microsoft PowerPoint) or clean manual schematics.

Thematic Reporting and Academic Formatting Standards

• Illustrative vs. Analytical Writing Modalities:

  • Module parameters prescribe an illustrative account over a purely abstract analytical account, emphasizing direct, grounded descriptions of empirical phenomena.
  • The report must be driven by authorial narrative synthesis; avoid stringing disconnected quotations together.
  • Exemplar Analysis from Terry (2016): Demonstrates effective illustrative reporting in a study investigating British male military veterans transitioning into civilian employment. The narrative utilizes the overarching motif of "translation," showing how participants like Brian experienced culture shock and systematically adapted their military communication styles to make them "civilian friendly," asserting a self-perceived inherent superiority of their original military-honed training discipline.

• Theme Selection Diagnostic Matrix:

  • To select the top $2$ to $3$ themes for the final report, evaluate candidate themes against four criteria:
  • Evidentiary Backing: Are there unambiguous, clear participant quotes to support the theme?
  • Internal Dichotomy: Does the theme incorporate internal tension, dualities, or opposing experiential viewpoints?
  • Analytical Depth: Does the theme transcend superficial topic summaries of interview questions?
  • Explanatory Power: Does the theme provide a compelling, nuanced answer to the research question?

• APA Quotation Formatting Protocol:

  • Short Quotations (Fewer than $40\,\text{words}$): Embedded seamlessly within the running paragraph, enclosed in standard double quotation marks, followed by the participant's designated pseudonym in parentheses:
  • Example: Effective teams can be difficult to describe because "high performance along one domain does not translate to high performance along another" (Ervin).
  • Block Quotations ($40\,\text{words}$ or More): Positioned on a new line without quotation marks, indented exactly $1.27\,\text{cm}$ from the left margin (Home > Paragraph > Paragraph settings > Indents and spacing > Indentations > Left: 1.27 cm). The pseudonym is placed at the end:
  • Implementation Rule: Used sparingly when extensive continuous participant narrative is essential.
  • Multi-Speaker Interview Quotations: Employs a hanging indent of $1.27\,\text{cm}$ (Home > Paragraph > Paragraph settings > Indents and spacing > Indentations > Special: Hanging > By: 1.27 cm), explicitly indicating speaker labels:
  • Example Structure:       Interviewer: Why did you decide to start using AI?       Interviewee: Well, my friends had been using it to help with their coursework and I was curious to see how it worked.
  • Quotation Density Rule: Maintain a balanced ratio of approximately $1$ supporting quotation per conceptual claim.

Student Experiences of AI: Empirical Findings

• Theme 1: Cognitive Scaffolding, Academic Utility, and Technical Failure:

  • Demystifying Academic Jargon: Undergraduates heavily utilize Large Language Models (LLMs) like ChatGPT as explanatory scaffolds to parse intimidating, jargon-heavy scientific literature, particularly dense neuroscience research containing specialized acronyms. Students prompt models to explain complex theoretical paradigms in concise terms (e.g., $20$ to $30\,\text{words}$), lowering academic anxiety and making research accessible compared to first-year baseline studies.
  • Ideation and Overcoming Cognitive Blocks: LLMs serve as dynamic sounding boards to resolve writer's block during coursework preparation. Students prompt the tool iteratively to suggest logical subsequent argumentative steps, generate essay outlines, or prompt revision questions.
  • Professional Communication Refinement: Students utilize AI tools to improve email etiquette, turning disjointed sentences into succinct, professional communications.
  • Technical Fragility and Hallucination: LLMs frequently exhibit severe functional limitations. ChatGPT displays profound computational incompetence with basic mathematical ratios and scaling (e.g., failing repeatedly to rescale a drinks recipe for $x$ individuals). Furthermore, free versions lack real-time access to scholarly repositories like Google Scholar, frequently fabricating academic citations and studies that do not exist.
  • Pedagogical Parallels: Students conceptualize the emergence of generative AI not as an uncritical replacement for intellectual labor, but as a digital study aid analogous to the launch of Google search engines.

• Theme 2: Surveillance Capitalism, Algorithmic Intrusion, and Digital Detachment:

  • Pervasive Surveillance and Ambiguous Consent: The integration of automated algorithms on social media (such as TikTok, Snapchat, and Instagram) generates feelings of invasive surveillance. Cross-platform tracking and predictive targeted advertisements are experienced as unsettlingly stalker-like. Students note the ethical ambiguity surrounding informed consent, as algorithms passively harvest user behavioral telemetry (e.g., watch duration, content saves, interaction frequencies) without conscious input.
  • Visual Skepticism and the Evolution of Unreality: Direct exposure to AI deepfakes, synthetic celebrity imagery, and generative voice cloning has altered student engagement with digital media. Students conceptualize AI image generation as an evolutionary continuation of historical celebrity Photoshop practices (e.g., the Kardashians). This normalization has resulted in profound digital detachment, cynicism, and cognitive dissonance, with students adopting a stance that no online content can be presumed authentic or anchored in reality.

• Theme 3: Academic Integrity, Institutional Panic, and Changing Pedagogy:

  • Threats to Authentic Coursework: Concerns exist regarding whether traditional coursework and long-form essays can survive, given the difficulty of distinguishing human prose from AI-generated text without robust detection software. Creative, humanities-leaning students express distress over the potential elimination of coursework, which provides space to demonstrate diverse competencies.
  • Institutional AI Infallibility and False Accusations: Irony exists in higher education institutions deploying algorithmic detectors like Turnitin to identify academic misconduct, as these automated tools can falsely flag authentic student submissions as AI-generated text, creating systemic institutional vulnerability.
  • Ethical Safeguards: Conscientious students establish rigid self-imposed boundaries: never copying and pasting generated text directly, generating authentic prose independently, using AI strictly for high-level structure, and cross-referencing every proposed factual claim and literature citation against empirical research databases.
  • Curricular Inclusion: Students advocate that universities proactively teach prompt literacy and ethical AI utilization within the formal curriculum, ensuring uniform competence across the student cohort.

• Theme 4: Existential Uncertainty, Human Agency, and Cultural Devaluation:

  • Labor Displacement and Existential Dread: Students experience anxiety regarding professional automation, fearing their prospective post-graduate job markets will be usurped by low-maintenance, cost-effective artificial systems, rendering human contributions obsolete.
  • Erosion of Human Creative Uniqueness: The automation of creative arts (music composition, illustration, painting) is experienced as an existential loss of human identity. Historical masterpieces were revered as the pinnacle of human expression; generative tools synthesize comparable outputs instantly, eroding human creative distinctiveness.
  • The Drive to Prove Human Efficiency: Some students deliberately resist deploying AI for coursework or daily domestic tasks (such as using Amazon Alexa to play music), motivated by an urge to prove their own cognitive efficiency and human value against computerized shortcuts.
  • Linguistic Courtesy and Anthropomorphic Masking: Students routinely embed interpersonal politeness markers (e.g., "please" and "thank you") when prompting LLMs, rationalizing that courteous prompts yield descriptive, nuanced responses, while humorously maintaining conversational courtesy as a safeguard against speculative machine dominance.

Questions and Interview Dialogue

Interview Transcript 2: Jess and Undergraduate Psychology Student

• Interview Metadata:

  • Recording Date: $19\text{th October } 2023$.
  • Audio File Duration: $00:48:58$.
  • Interviewer: Jess (Staff member, Newcastle University, Stage 1 Research Methods).
  • Respondent: Final Year ($\text{Stage } 3$) Undergraduate Psychology Student.

• Transcript Exchanges:

  • Interviewer: "So just to make sure that you're happy with what's going on, erm, do you consent for me to audio record and video record, erm, this interview?"
  • Respondent: "Yes."
  • Interviewer: "Erm, do you understand what this interview is about and you're happy to continue?"
  • Respondent: "Yeah, I understand and I'm happy."
  • Interviewer: "Okay. And you're happy for this, er, the transcript from this interview to be used in teaching?"
  • Respondent: "That's fine, yeah."
  • Interviewer: "Okay, perfect. Erm, so just for the recording, erm, I'm going to introduce myself and then for you to introduce yourself afterwards. Er, so my name's Jess. I'm a staff member at Newcastle University, teaching on the research method stage one module. Erm, and this interview's being carried out to create materials for, er, that module."
  • Respondent: "Erm, my name's [Name]. I'm an undergraduate third year psychology student at Newcastle."
  • Interviewer: "Perfect. So, [Name], when did you first learn about AI?"
  • Respondent: "Erm, I'd say when it all, it all started taking off, erm, things like Chat GPT and things like that. Er, maybe also as well when there started to be things like, erm, eye-tracking and things like that, erm, social media. I suppose social media was probably the first time that I'd heard of the use of AI."
  • Interviewer: "So what was it on social media that you were seeing?"
  • Respondent: "Erm, so I was seeing that things that you would, things that you would interact with or like, you then start seeing more of those things, erm, especially to do with like adverts and stuff as well. Erm, you could sort of tell that they were targeted specifically at you and your interests. Erm, yeah, I noticed that."
  • Interviewer: "Okay. Erm, so you've mentioned Chat GTP before, erm, so obviously that's a large language model and it can be used for basically anything. When you started to learn about those types of AIs, what was your initial thoughts about them?"
  • Respondent: "Erm, I think initially I was quite, I was quite excited about it. I wanted to find out what, like about it, what it did. Erm, so I did, I did like log in and everything and have a look at Chat GPT and have a play around with it. Erm, I was a bit sceptical about, erm, people started talking about whether it could replace, replace jobs, erm, things like that. Erm, so I think that did scare me a little bit. Erm, also, sometimes it can be so similar to almost a human interaction. I mean you can tell. I suppose it depends. I mean I was just using the free version but I'm sure there are more advanced ones that I could imagine are quite similar to human interactions. I think it would be difficult to tell, erm, maybe going forward, whether you were talking to a person or a computer maybe. Erm, so I was a bit, er, scared about that. Erm, also, it was mentioned that maybe coursework might just be completely like dumped or like gotten rid of because how do you, like how do you know if it's an AI, erm, created essay or human created one? Like I understand it's difficult to, erm, to like, you know, without, so if you didn't have software that could detect that, erm, that it would be difficult to work out whether it is really student work that you've got."
  • Interviewer: "So how did finding out that this could remove coursework from your degree, how did that make you initially feel?"
  • Respondent: "Erm, I wouldn't say I was happy about it. [Laughter] Erm, I did think, you know, obviously I'll believe it when I see it but, erm, I'm glad that now it seems to have starting to go more towards, er, sort of… I've heard people compare it to when Google came out, that, you know, sort of you can use it more as like a study aid rather than a replacement. And obviously like I said, it seems like we do have software that can work out whether a student has written it themselves or not. I know I have, I have used, erm, Chat GPT for my studies as well and I have found it useful."
  • Interviewer: "Yeah. And say before, you just mentioned that when you first heard, it made you like a bit upset about the loss of coursework. Why was that?"
  • Respondent: "Erm, I'm quite like, I've always been more of an English person than a maths person. I like, I really enjoy like the creative side of it and I wouldn't want that, erm… I think it, it would change courses. Not just psychology but I think it would change courses quite significantly to just not have coursework anymore. Erm, I think it gives a chance to show different sets of skills. I think yeah, it would be a shame to not include that still."
  • Interviewer: "Yeah, no, yeah I understand. Erm, so what about coursework changing for the inclusion of AI? How would you feel about that or do you think that—"
  • Respondent: "I think, when you say change?"
  • Interviewer: "So some of the things people have talked about is having it so that, you know, you actively ask students to use AI and teach them how to use it and how to use it effectively for their, for their coursework."
  • Respondent: "Erm, I can't see that being necessarily a bad thing. Erm, I mean I have used it for, for things myself before. I think it's quite good at, erm, it's quite good at clarifying, like clarifying quite complicated, you know, pieces of research and theories and things like that. Erm, I think—"
  • Interviewer: "Sorry."
  • Respondent: "Sorry. Erm, I was just gonna say I know that, erm, I've spoken to my friends about it and we agreed that it does, when you hit a mental block, it can just help you just to get started again, just to give you… I mean it can give you way too much information as well so I think you'd have to be quite specific with it and say oh, could you, erm, explain this theory to me in like 20 or 30 words. Erm, but I think it can be really, it can be really useful for that. I think teaching other people, teaching everyone so we're all sort of on the same page so no one's really at a disadvantage, I could see that, I don't see that as a bad thing."
  • Interviewer: "So I know you just gave me examples saying like you give it… sorry, could you just repeat that example again for me that you gave?"
  • Respondent: "About like it clarifying quite complex—"
  • Interviewer: "Yeah. So an example of when you've used that."
  • Respondent: "Erm, I'm trying to think. When have I used, so … if I come across a study say, erm, just say for like neuroscience, neuroscience sort of based study and it's got very specific, particular language, many, many acronyms and it's, it's very jargon-y, I think, I know I've highlighted like a passage and I've put it into Chat GPT and said oh, could you simplify this for me. Could you explain this part? I think it's really good for that because, especially when you're in the, like your first year of uni or in second year and it's quite, looking at all of these, erm, yeah, quite, like in-depth research studies, it can be a bit daunting, erm, with all the jargon. And I think Chat GPT is really good for simplifying it cos I mean, you can search in Google obviously, in, you know, things like simple psychology but they only go so far. Whereas Chat GPT can be, and other AI things like that can be more specific to like your needs and what you're looking for."
  • Interviewer: "So obviously you've already made the comparison to Google, so have you got an example of where you've tried to find something in Google and then you've ended up using Chat GTP or have you found that is Chat GTP always better than Google have you found then?"
  • Respondent: "I wouldn't say it's always better than Google, no. Erm, for example, the free version that I've used, it doesn't have access to Google Scholar. So if you ask it about, I don't know, erm, Bander et al 2000 and whatever, erm, it doesn't, it can't necessarily tell you about it. Erm, it will come up with a message like I don't have access to any articles or anything. And it also doesn't always have the most up to date information. I know I've searched things before and it said I only have access to records, I don't know, since whatever date. Erm, so I definitely don't think it is always as good as Google but saying that, I know one of my friends who does computer science, they paid for the more advanced Chat GPT and they said it, it does make a difference, it is more accurate than like the free version. It does… yeah, they found it, they found it really useful. I don't know if that would include like more access to things like articles and stuff but yeah."
  • Interviewer: "Yeah. So I know you're saying that it's not always the best thing to be using but when you first went to use it, what made it appealing to you to begin with?"
  • Respondent: "Erm I mean at the start I didn't really, I didn't really have many like pre-conceptions cos I was not really sure like how it would work and then when I started using it, I realised it was sort of like you know you type in like a request and then it would give you the answer back. Like I did think that was really good but I did notice that it wasn't, like I said, it wasn't always accurate. And sometimes I'd have to say, 'Are you sure?' like it's almost like human interaction but I, it's almost, almost like a text, like the, write a text back are you sure about that. Erm, and then it would say oh sorry, I got that wrong but it's actually this. And I, I realised that if I said that like, like sev-, like three times, it would constantly say oh sorry about that, I got it wrong. Erm, so I think that is a bit problematic. It was quite a simple thing I was asking it, like maths related. It was to do with measurements [I: yeah] and it kept giving me wrong answers. And I was thinking if it can't get that right, I'm a bit worried about what it's told me previously [laughter]."
  • Interviewer: "So knowing that it gives you the wrong answers, has your approach to it changed?"
  • Respondent: "Yeah, definitely. Since, like when I realised that it had been, er, it it can give you wrong answers, I think it makes you realise you do have to be, use it very carefully and make sure that, er, you know, what it's, what it's saying, you know, you think is definitely right."
  • Interviewer: "And so what, what steps do you take to make sure that what you're getting out of it is correct?"
  • Respondent: "Well in the, in the, so in the example, it was a recipe and I was trying to make it for, it was, it was a drink and I was trying to make it for a lot more, a bigger quantity than what I had. And, erm, yeah, like the measurements of the different ingredients needed to be right and I just, in the end, just worked it out myself because it would say like could you, I gave it like a res, like an ingredients list and a method and I said could you make this so it was like for, erm, I don't know how many people but just for x many people. Erm, and yeah, once I realised that it just couldn't seem to, it just couldn't seem to deal with the like ratios and things like [I: mm] that so I just ended up just calculating it myself, which is what I probably should have done in the first place but I was being lazy."
  • Interviewer: "Bless you. Erm, so can you remember the first time you used AI? What did you use it for?"
  • Respondent: "Erm, that's a good question. I think it was to, I think I asked it to summarise. like summarise, erm, like pieces of research, like complicated pieces of research and, erm, yeah, define, like defining things as well."
  • Interviewer: "So like the example you gave before where you'd give it a paragraph and—"
  • Respondent: "Yeah."
  • Interviewer: "Okay. So what caused you to try it that first time?"
  • Respondent: "Erm, I think I was just, I was interested in, in how it worked and what it did. I'd heard, I'd been like recommended to use it by a friend, and they'd said that they found it really useful and that I should, I should try it. Erm, so I think I was kind of like motivated, motivated to try it because it sounded like, you know, it's like well they were using it and it was good and they recommended it. So I suppose that's what sort of motivated me to start."
  • Interviewer: "and have you use, I know you've mentioned using it for the drinks recipe and for summarising like, like journal text and definitions. Have you used it for anything else? I know we've mainly been talking about Chat GTP but any other AIs that you've used as well?"
  • Respondent: "Erm, so I've definitely used it for, for recipes, like food or drink. Erm, I've used it for like summarising things or clarifying things to do with like my course, to do with psychology. Erm, as I mentioned the articles. Erm, what else have I used it for? I've used it for trying to, not so much recently but I know to start with I did. I did use it to make … things like emails sound a bit more professional, especially if it was about something important. [I: yeah] Erm, I'd just put it through Chat GPT and say oh could you make this more professional or could you maybe make it a bit shorter and stuff. I mean I feel like I am okay at that anyway. I just, it's just nice to have like, almost like a second opinion. [I: yeah] And yeah, I'd say sometimes that's helpful but sometimes it does just give you so much, so much text and so much information and it makes it too long. So I'd have to say like oh you know, can you make it x words, [I: yeah] erm, to try and shorten it down but I think it is quite good for that. Erm, I know one of my friends is more of a, they're really good at maths but English is not really their thing [I: mm] and I think they found it really useful for things like that. Writing like a professional email, erm, is not really something you ever get taught, necessarily. So I think they, I know that they found it really useful and like I said I did, I do as well."
  • Interviewer: "Just using that example of using it to improve like your email etiquette, erm have you taken, is there anything that Chat GTP suggested that you've then taken into when you write an email?"
  • Respondent: "Erm, it's hard to think of specific examples but … I think just the wording, like the wording used to make it sound more professional. I mean obviously you do have to make sure it does still sound human and not like you have just asked Chat GPT and copy and pasted it. Erm. … I think it can, it can explain things quite like eloquently. It can clarify, erm, maybe say you've written it in two different, two or three sentences and it can just sort of give you one sentence that seems to pack it all in there and sound quite good. So I can't think of like a specific sentence but I know that that's the sort of thing that I would have used, rather than copying and pasting a whole, almost essay worth of a, an email that its output."
  • Interviewer: "Yeah. And say, I'm trying to think, so you mentioned you've used it a few different ways. Erm, has it always been used in relation to, I know you gave the drinks recipe example but is it, do you mainly use it for university work?"
  • Respondent: "Erm I'd say it's pretty equal between erm university work and erm sort of recipe, recipe things. Like sometimes we don't know what to make for dinner and it's quite good, you just ask, ask Chat GPT. I know I saw a Tik Tok where a girl had planned her whole day using AI. She'd asked like could you plan me, you know, make my weekend plans for me and it had output her a, like a plan of what, what to do. And I think some people seem to find that quite controversial. I like, you know, it seemed like she, you know she enjoyed it and it was like, it did give her quite a good, erm, a good plan. But I think some people found it a bit like why are you letting AI dictate what you do? Can you not, you know, (a) can you not just think for yourself and (b) you're kind of letting a robot almost tell you what to do."
  • Interviewer: "Yeah. And so would you ever try that yourself?"
  • Respondent: "Erm, I think maybe yeah. I feel like I would definitely ask it things, like ideas of things to do if I was stuck for ideas. Like I said before, I think it definitely does help you get past like a mental block [I: mm] cos it can just, even if it just gives you something that just starts, starts you off thinking again. Erm, yeah, I think that can be really useful."
  • Interviewer: "Have you ever used it for like, I don't know, well for writing coursework?"
  • Respondent: "Erm, I have, I have, yeah, I have used it during coursework, yeah."
  • Interviewer: "I'm just trying to think, when you're using it, what was kind of going through your mind?"
  • Respondent: "Erm, I think I was just thinking, so say it was for like, erm, an essay or something and you've like got a couple of, you've got a couple of like quite good points already or a good point and then you're sort of thinking like what's, what's the next one, erm, I think it can be quite good for just giving you a little, you know, asking it information about the topic and then it can just give you a little jolt to think, 'Oh, that can be my next point.' Or you ask it about some research and then you sort of think oh, like the way that it's, it's talked about it sort of can help you piece it, like can help me piece it together. Erm, I mean I wouldn't say I've used it like exhaustively. It's just it's more if I'm, if I'm stuck and I'm struggling to get past something, a problem."
  • Interviewer: "That block that you keep mentioning."
  • Respondent: "Yeah. Or just even like, yeah, just… it's a, it's almost like a problem, like a problem solver. It can get you past like, you don't understand something or if you can't think, you're struggling to think of something new and stuff like that."
  • Interviewer: "Yeah. And so when you say that you use it to get you past that block and using it to identify like another point, how are you getting it to give you that like additional point for your essay?"
  • Respondent: "Erm … I think I would just… I mean it's quite good because you can just ask it question after question after question. You can say like could you expand more on this part that you said or this part that you said. Erm, I think yeah, that's sort of the way I would go about it."
  • Interviewer: "Okay. So would you ever feed AI any information or do you just take the information that it gives?"
  • Respondent: "Obviously you have to feed it information for it to give you, for it to give you the information back. So I would have to tell it things for it to."
  • Interviewer: "Okay. So when we're talking about like getting this next point, what information would you give it? So would you give it like information from papers or, you know, like ideas that are already in your head or how would you—"
  • Respondent: "Yeah. I'd say both. I would give it sort of maybe what I'd said before, erm, yeah, maybe some ideas that I had, sort of like the rough, the rough like topic area that the thing is about cos I think that is something it can struggle with as well, is what… like it needs context. It can't like give you, sort of answers without, it can't, it can't give you any answers without any context but if you don't give it the right sort of context, it doesn't even know like what sort of essay you're writing. [I: yeah] So it could give you something for… yeah, it doesn't quite, it doesn't work. So yeah, definitely ideas that are in my head, like maybe some research articles that I was thinking of using, the sorts of information that are in those. I mean obviously there is a line and it, you know, you can't, it is plagiarism. Erm, so I do try and avoid that at all costs. But obviously it's quite a new, it's quite a new recent thing. Erm, so yeah, I mean there's not been, not been too much opportunity to plagiarise so far."
  • Interviewer: "So what steps do you take to avoid plagiarism when you're using these AI models?"
  • Respondent: "Erm, so, well I always just, I always write the essays myself. I would never just copy and paste the output into my essay. You know, I have to make sure the ideas are my own. I mean obviously as you said, it kind of gives you a jolt but I think there is a diff-, I mean obviously there is a difference between AI sort of leading you in the right direction and then as opposed to just going like oh, it's given me all the answers, whack that in there. [Laughter] Erm, so yeah, I just try to make sure that, I mean obviously make sure that all my points are backed up with research. I think that's the main thing, if, if I can back it up. If I can back up these points with research, then I don't see really too much of a problem."
  • Interviewer: "So when the AI gives you kind of your next essay point, erm, to back it up with research, do you get the AI to tell you where the research is or do you then go and search for that yourself?"
  • Respondent: "Erm, no, I would always search for it myself. Erm, I actually don't think that it can do that. Like I said, I don't think I even has access to. Erm, I have asked it for research before. I've thought of a point and then I've really struggled to find research to back up my point [I: yeah) even though I kind of know it is, it is like the right sort of thing. Erm I really struggle to find any research to back it up and then I'll ask, so I'll ask Chat GPT or an AI model to, erm, do you know any research that shows x. Erm, but it has suggested articles to me and then I've looked them up and they just don't seem to exist. So I'm not sure where, which records it has access to."
  • Interviewer: "Yeah."
  • Respondent: "Like either the, not ones that we have access to or they, it is just sort of giving you something that doesn't really exist, I don't know. Erm, but I know I definitely had that before where yeah, it is output studies that I can't find, no matter how hard I try."
  • Interviewer: "So have you had any other kind of, erm, instances with Chat GTP where maybe it's given you false information or, cos you mentioned about it not being able to figure out the measurements properly and about these references maybe not existing. Was there anything else you've experienced?"
  • Respondent: "I can't think, I can't think, I don't use it that often."
  • Interviewer: "Okay."
  • Respondent: "Like I use it, I do use it but I don't, I can't think of anything else where it is output like wrong information and I feel like, I mean hopefully I have realised if it has. Erm, but it's definitely something to, to keep in mind."
  • Interviewer: "So have you ever thought about what it brings out critically before?"
  • Respondent: "Erm, do you mean like thinking about being critical of what's—"
  • Interviewer: "So being critical of what it's saying and checking that what it is saying is correct."
  • Respondent: "Erm, yeah. So I think, I mean especially since sort of clocking that it's not always spot on and sometimes quite out, quite wrong as well, I have, I am definitely more critical of it cos I think it's quite easy to be quite trusting. Erm, yeah, it's quite easy to be quite trusting. I feel like when you search… it's different to search engines as well because when you're searching Google, erm, it gives you sort of, it often will like put in bold or slightly bigger, the answer that it thinks that you're looking like at the top or [I: yeah] sometimes it has the options of the drop down arrows and you can like go down and it's sort of very similar questions that people have asked. And usually you can sort of tell that they all are overlapped and they all sound like the same sort of thing worded slightly differently, especially then if you like go on to a website and you check or if you go on to, you know, any, you know, just online, you know, in books or anything like that and you think yeah, that seems like the general consensus [I: yeah] so. But with Chat GPT, I think it's quite easy to just ask it, like similar to Google, ask it a question and then trust that the output is correct. Erm, and there isn't, there isn't really a way on, I mean I don't know about others but on the AI that I've used, there isn't really a way, without just saying back to it are you sure, and then it comes back with, sometimes, yes or oh actually, on reflection, I've realised that this is incorrect information. Erm, so yeah, I think that's definitely a danger of it."
  • Interviewer: "And so other than asking it if it's sure, is there any other steps that you take to check what it's telling you is correct?"
  • Respondent: "Yeah. I mean I would always, I would always just go and check through. I mean if it was, say with the the measurements, just common sense sort of said to me that that doesn't sound right, you know, that doesn't seem right and those sort of numbers are not what I'd expect. So I suppose I would compare it to my expectations of what I was going to see in the output. Erm, I'd compare it to, if it was more research based, I'd compare it to sort of oth-, like research articles and things like that. Erm, I think there is always a way to check it. It's just remembering to check it, remembering to think to yourself, 'Is this, is this correct?' Erm, but I think yeah, that is, I mean I'm sure this is not just, erm, like a personal thing. I'm sure that other people have definitely, using it as well, have realised this and realised that then okay, I need to be more critical when I'm using it."
  • Interviewer: "Yeah. And so have you have got an example, I know you gave the drinks recipe one, but when you've been asking it information about an essay? Have you got an example of where it's been completely out?"
  • Respondent: "I'm trying to think. Erm, been completely out? Erm."
  • Interviewer: "It's okay if you can't think of one."
  • Respondent: "I'm trying to remember what, what I've asked it and what it… erm, I think, I can't remember what it was but I was asking it for a definition and it did give me an incorrect definition. [I: mm] Erm, and I think it was, I can't remember what it was a definition of but I remember I asked it for a definition of a psychology term and it output something completely wrong. Erm, I mean you can report it on there as sort of like as was it a helpful response. You'd say no."
  • Interviewer: "Do you often, er, give feedback as to whether or not it was a helpful response or?"
  • Respondent: "I did more to start with. I don't seem to, I don't so much anymore. Erm."
  • Interviewer: "How come?"
  • Respondent: "I think just, I don't know if you remember when like the Amazon Alexas came out and at the start everyone used to say like please and thank you to them. And then the more they had them, the less they did it. Erm, but weirdly enough I do, I do when I'm asking AI a question, I do say oh please can you tell me this and thank you afterwards because, partly because like obviously it is nice but partly, erm, because it gives, well I don't know if this is true but I think it might be, it can give you sort of, more descriptive and better responses if you are courteous to it. I don't know if that's 100% true but I'm gonna do it anyway in case robots take over the world. [laughter]"
  • Interviewer: "So what led you to like, like what sort of, well what led you to thinking that adding please and thank you led to better outputs? Did you see it yourself or?"
  • Respondent: "Erm, I can't remember whether I saw it myself. I know I spoke to someone else about it and I think we, yeah, I think I spoke to someone else about it. I can't remember if they, I think they might have said it to me. Er, but I definitely notice just, I mean it might not be to do with courtesy but I noticed that if I am more, if I give it a shorter question, it gives me a shorter, a shorter answer. So if I say like, I don't know, a question that is much more blunt, it gives me a much blunter answer."
  • Interviewer: "Okay."
  • Respondent: "Erm, so I don't know if by saying like please and thank you, maybe that's sort of like, cos it's lengthening it, it then gives me a lengthier answer maybe. I don't know if that's got anything to do with it. Erm, I mean I guess it would probably, I mean it is AI so it is looking at what I'm saying to it and learning, you know, from my like,"
  • Interviewer: "yeah"
  • Respondent: "just learning from me and that's changing its behaviour constantly so."
  • Interviewer: "So you mentioned a few times that your friends have used AI. So do you know what types of things your friends are using it for?"
  • Respondent: "Erm, I think, I know, I mean a lot of my friends do different courses and I know that, erm, the one that does computer science was using it for, er, computer, computer code output or like he was asking it to make a code."
  • Interviewer: "Yeah."
  • Respondent: "Erm, say like Java or something. And then using that when, in like times where he was maybe not so sure and stuff like that. Erm, I think generally my friends do use it for course related like things. Erm, yeah, I think that's fair."
  • Interviewer: "So have you or anyone you know ever used it for like exam prep?"
  • Respondent: "Erm, when you say exam prep, do you mean like revision techniques or?"
  • Interviewer: "Erm, it could be revision techniques, it could be prompting it to give you like potential exam questions, just anything."
  • Respondent: "I've used it that way myself actually and forgot but yeah, I have, I have asked it to come up with, erm, yeah, exam, like practice exam questions."
  • Interviewer: "And was that helpful at all?"
  • Respondent: "Erm, I think it was. Erm, I think it is obviously, it is a computer model. It doesn't know, it isn't, you know, it's not like it's, it's completely personal. It is just using the information that it has. Erm, but I think it was helpful in giving an idea of the sorts of questions. Erm, yeah, yeah. I think—"
  • Interviewer: "And so when you were prompting it to get those sorts of questions, what happened next for you?"
  • Respondent: "Erm, so I would, I would pick some of the questions and then like try and like write a, write a brief, you know, essay for them or just, even just looking at the questions, like could I answer that? Do I know what sort of things I should say? Erm, whether that be like an essay sort of style question or a short, even just a shorter question like do I know the answers to these. And then if there's some that I think I don't know what I'd write for that, going back and then revising that, those ones more. Erm, yeah."
  • Interviewer: "So to get those prompts, was that the same process as you did for essay writing or did you have a different process?"
  • Respondent: "When you say process?"
  • Interviewer: "So for the essay writing, you were saying that you would put information into it, erm, to then ask it to make the next point and things. Was it the same type of thing, were you putting information from lectures in or?"
  • Respondent: "Yeah. So if I, if I asked it to, to give me some, some questions, erm, and they weren't necessarily essay style, they were like shorter ones, erm, and then I wasn't sure of the answer, I might then ask it what, how would you answer this question? But then I would go back to like the lecture material and make sure that like I (a) I understand it, (b) it sort of aligns with what it said, erm, so yeah."
  • Interviewer: "So do you think you've had a positive or negative experience of AI like overall?"
  • Respondent: "I'd say positive, erm, in terms of me like using it for like my own, you know, interests, studies, things like that. But, erm, I think it can be a bit, it can feel a bit invasive when it's on things like social media and it's, erm, you know targeted. targeted ads. You know, not so on social media but say like using AI as tracking, you know, sort of targeting ads towards you, er, what you're looking at, you know when you've just looked at something and then as soon as you go on another website you see an advert for it, I think that is a bit, it's very invasive. It's, you've learnt, like you've already learnt from what I was just looking at, that that's something that I'm interested in and now you've put it on another website. It's almost a bit stalker-like. Erm, I think, I know on, erm, you know, back to the social media, like the algorithms, I think there is, I think I'm on the right lines saying that AI is involved in that, in like tracking what sort of things you're interested in. It can be useful because I mean, so for instance Tik Tok, that's an app with a lot of, a lot of users, a lot of different content and I know that my Tik Tok and the sort of content it gives me is completely different to say what one of my other friends, I, I see theirs is open next-, in front of me and the content that's on there is just completely different to what's on mine. And I wouldn't be interested in some of the things that they see because they are relevant to them. So I think in terms of that, it can be, that is useful but, erm, it can be, it can be quite invasive. I know, yeah, I don't know, it just can feel a bit almost like you've got a stalker that knows exactly what you're interested in, almost too specific sometimes. It's sort of a bit scary."
  • Interviewer: "And so you talk about it being very invasive and quite scary, erm, what specifically about the use of AI is making you feel that way?"
  • Respondent: "Erm, I think it's providing a lot of… it's taking a lot of information about you, [I: yeah] erm, and then it's obviously learning from that and then it's, you know, outputting things. Erm, I just think it's almost sometimes without even your, it's, you're not even always consciously aware. I mean I am aware that it's going on but you forget. [I: yeah] And then when you see something that is so specific, it's almost like they know you or it's almost like someone that you know very closely has shown that to you. I think that can feel quite erm, it just feels, it just feels unusual that, erm, a computer, like a computer knows your, say like your interests or something or just something about you so well that it would know to show you that. [I: yeah] It can just, yeah, it can just feel a bit like… it's hard as well to use these things without, you know, without sort of consenting to to that. And I think like the area of consent around AI is quite, erm, grey. [Laughter]"
  • Interviewer: "Yeah. And so obviously we're talking about two various different types of AI throughout this interview. We're talking about the large language models like Chat GTP and then we've been talking about algorithms as well."
  • Respondent: "Mm-hmm."
  • Interviewer: "Erm, do you feel like you have more sort of control over what happens with like the large language models compared to the algorithms through, erm, social media?"
  • Respondent: "Definitely because, erm, I know I'm consciously inputting into the large language models but with the algorithms, it's very passive, passively, erm, interacting, passively… I mean I say passive, I know some people do purposely interact with more, they interact with things because they want to see them again but I feel like for me, most of the time I'm just, I'm passively contributing to its knowledge just by, you know, what I'm saving, like which, you know, content I'm saving, content that I'm maybe watching for longer, if I rewatch it again or something. Erm, I'm not consciously thinking about oh I'm, I'm providing it with information, it's just happening."
  • Interviewer: "And so do you take any steps to try and like, to try and avoid the, I want to say the algorithm learning about you. I don't know if that's the right sort of description but."
  • Respondent: "Erm, I wouldn't say I take any steps to avoid, avoid it learning about me. Erm, I think, I don't know. Yeah, I don't, I wouldn't say I do, no."
  • Interviewer: "Okay. Erm, and so thinking back to, cos obviously a lot of it, I know tracking on social media and things has been going on for a long time but thinking back before like the use of AI or before AI was accessible really to the everyday person, like has your life changed much since then?"
  • Respondent: "Erm, as in like, when you say has my life changed much, do you mean because of the AI?"
  • Interviewer: "Yeah, because of AI."
  • Respondent: "Erm, that's tricky, a tricky question. Er, I'm trying to think how, how it would have, how it could have. Erm, I mean it's not something, I mean obviously before it was around, erm, before it was like as, I don't know what the word is but I don't want to say popular but I'll just say like available. Before it was so available, I would have just relied on things like Google, er, and books, things like that. So I suppose that has, that has changed in my life. Erm, but then I wouldn't say I rely on it. It's more, yeah. I don't, yeah, I can't, I'm not sure, sorry."
  • Interviewer: "That's fine. So AI came, I'm assuming, during your second year of university?"
  • Respondent: "Yeah."
  • Interviewer: "Is that right? So thinking about what happened in your first year when AI wasn't around to now, has much changed in how you use it for your university work? So—"
  • Respondent: "I think it's definitely made some, it's definitely made like jargon a lot, I feel like it's made me less wary of really jargon heavy like studies and research like that. Erm, but then I suppose just being exposed to that naturally through being at university might have made me less sort of, erm, I was going to say overwhelmed but that's not what I mean. Like, I don't know what I mean but erm, I think it's, it's made it more accessible for me. It's made the research studies more accessible for me because I don't feel intimidated, that's what I meant. [I:yeah} It's not made me feel so, erm, intimidated by language that I don't understand, erm, because I know that I can, I can look at this study and if I don't understand the language, I'm not going to be at a complete sort of (a) loss or (b) just like spending a lot, a lot of time trying to understand it because I can just put it into say like Chat GPT and it will sort of, it can sort of summarise things and sort of help me, help me along with those sort of things. That's definitely not something I had in first year. I definitely, I definitely did struggle with that a little bit. Erm, so I would say it's made my life easier in that sense."
  • Interviewer: "Do you ever feel like you may have become reliant on it or?"
  • Respondent: "I wouldn't say reliant, no. Erm, I feel like it gives me a different way to get information and like sort of help come up with new ideas and things. Erm, I know I've been helping plan an event for, erm, like a science event for children recently and I did ask it sort of to help me come up with some ideas for, I'd sort of been thinking for a while of different activities and things to do. And I thought well I just, I can ask and see if it comes up with anything different, any new ideas. Erm, I don't think, in the end, I went ahead with, with anything it, erm, it suggested cos I think often it suggests things that needed quite a lot of resources. [I: yeah] Erm, but it was useful. It was useful to, to have that. I wouldn't say I was reliant on it but it was, it was a useful tool to see what it came out with."
  • Interviewer: "Yeah, so using it as a bit more of a sounding board rather than as like a, getting it to do everything for you?"
  • Respondent: "I'd say yeah, I'd use it as, as like an aid, not something that I would like, I would like have to have."
  • Interviewer: "Yeah. And so overall, do you think, like towards your university work, do you think AI's been a positive then as a student?"
  • Respondent: "Erm, I would say it's been a positive, yeah, because it is, it's helped me to, you know, feel less intimidated by like the language I was saying about. Erm, it's, yeah, it's helped, yeah, it helps to clarify things. I would say it has been helpful overall as a student, erm, but obviously with a pinch of salt."
  • Interviewer: "Yeah. And so do you think it's had any negative impacts on your ability to learn at university?"
  • Respondent: "Erm, when you say negative impacts on my ability to learn, what do you sort of mean?"
  • Interviewer: "I mean just sort of like any sort of negative impacts on just like your university experience or anything that's happened like within the university sphere of your degree. Well I say your degree but just like university in general basically."
  • Respondent: "Erm, I don't, I wouldn't say it's had a negative impact, no. I mean like I said, it can be, it can be wrong, erm, which is obviously not very useful. Erm, so I suppose that is a negative that, you know, I didn't necessarily initially realise that it could be so incorrect sort of and not even realise itself that it was incorrect until you'd asked it like are you sure about this. Erm, so I would say that is a negative that I could have potentially have asked it information, received incorrect information back, erm, which is obviously negative for a learning experience and like ability to learn. Erm, but yeah, I don't think it's necessarily affected my, I mean I just said that but I don't think it's necessarily affected my ability to learn. I don't think it's changed that in any way. It's just, erm, yeah, I'd say if anything, it has encouraged me to be more critical of the information that I've received so that is almost, you know, a benefit."
  • Interviewer: "Yeah, definitely. And so going forward, I know you're in your final year now, erm, but looking forward to what the plans you have next, so maybe master's, maybe like a job, can you see it having any benefits going forward, you having learnt to use it at university?"
  • Respondent: "Erm, I think, I think I would definitely still use it in similar ways, like continue using it in similar ways to how I have already. Erm, the more like I learn about it, I suppose there is the potential to use it in, in different ways. Erm, I think yeah, I think, erm, yeah."
  • Interviewer: "That's okay. We're getting into the harder questions. Erm, and so thank you for your time. Erm, is there anything that you want to add that maybe we haven't covered or you just want to say?"
  • Respondent: "Erm, I think that is, erm, that is everything. I can't think of anything else off the top of my head."
  • Interviewer: "That's fine. Thank you very much for your time. I'm going to stop the recording now."
  • Respondent: "Thank you."

Interview Transcript 1: Gary and Undergraduate Student

• Interview Metadata:

  • Recording Date: $19\text{th October } 2023$.
  • Audio File Duration: $00:34:04$.
  • Interviewer: Gary (Interviewer, using a Dictaphone and prompter screen).
  • Respondent: Undergraduate Student (Participant 1, Transcript 1).

• Transcript Exchanges:

  • Interviewer: "Okay, happy with that? Have you got that your end?"
  • Respondent: "Yes. Yeah, I got a like a little notification as well saying like it's being recorded so I got it on my end as well."
  • Interviewer: "Perfect. Okay, well we'll, er, we'll get started. Good morning, [Name]. Erm, my name's Gary. Are you, er, are you happy for me to call you [Name]?"
  • Respondent: "Yes, [Name]'s fine, yeah."
  • Interviewer: "Okay, perfect. So thanks very much for joining me today and agreeing to, er, to help with this interview. Erm, just so that you know, I've got another screen with me and I'm gonna sort of occasionally look across to the screen, erm, for prompts that I, erm, that I need to, to make sure that we, when we're talking that I don't miss any information out. Is that okay?"
  • Respondent: "Yeah, that's okay."
  • Interviewer: "Right. Erm, so I think you've, er, you've agreed to help with this study which is what is the undergraduate student experience of AI, artificial intelligence. Erm, can I just confirm that as far as you're concerned that's, that's what you've been involved with."
  • Respondent: "Yeah, that's all correct, yeah."
  • Interviewer: "Oh right, perfect. So what we're gonna do is we're just gonna go through an interview process today, erm, where I'm just gonna ask you a few questions, er, about that. Erm, please take your time and give as much information as, er, as you feel able to, erm, about that as we go on and we'll just, er, we'll just take it as, as we go. So, erm, there's a certain number of topics and questions that I need to go through but don't feel constrained, erm, by what you want to say so please just, just say it and if, erm, as we go along, I'll guide you towards the right, the areas that we're interested in talking about, okay."
  • Respondent: "Yeah."
  • Interviewer: "Excellent. So, erm, can I just confirm that you've had information on the, erm, interview this morning?"
  • Respondent: "Yes, I have."
  • Interviewer: "Okay, great. Now, one of the things, erm, that I'd like to know, is there anything that's likely to affect, erm, the interview taking place? Is there anything I need to know that would make it difficult for the interview or are you in a, erm, a comfortable place? You're not gonna be disturbed? Have you got limited time, that kind of thing?"
  • Respondent: "Er, no, there's no like limitations. Everything should be fine."
  • Interviewer: "Okay, great. Thanks very much. Erm, there's a couple of things that, towards the end, I mean firstly, before we start the interview, erm, I'm gonna need to, to check the, er, the consent with you, that you're okay for the interview to be, erm, to go ahead and that, erm, you're okay for it to be recorded. And at the end of the interview, erm, I'll arrange to send through a, erm, more further information in a debrief form at the end, okay."
  • Respondent: "Mmm."
  • Interviewer: "So first of all, erm, the interview itself is probably likely to take between half an hour and 40 minutes, okay. And, erm, can I just, er, confirm that you're happy for the interview to be recorded?"
  • Respondent: "Yes, yeah, I am."
  • Interviewer: "Okay, great. Thanks very much. And you understand that the interview recording will only be used as, erm, as teaching materials and for no other purpose?"
  • Respondent: "Yeah."
  • Interviewer: "Okay, that's great. Have you got any questions on any of that?"
  • Respondent: "Er, no, no questions."
  • Interviewer: "Okay, lovely. In that case I'm gonna turn on the Dictaphone, erm, to make sure that, er, that we can record the interview and, erm, we can start the interview as, er, as we go on. So the first thing, erm, I'd say as well, just before we start, is that I might look away because I might be taking some notes but it doesn't mean I'm not listening to you, okay. I know it's a little bit difficult when we're online cos, er, as soon as I'm not looking up then it might look as if I'm not paying attention but I will be, okay."
  • Respondent: "Yeah."
  • Interviewer: "So first of all, can I ask you, where did you first, er, hear about AI?"
  • Respondent: "Erm, probably around like movies and TV shows, kind of that's probably where I've probably heard about AI and like it's kind of been mentioned kind of like years before it even became like a real thing, like in popular culture, in like articles kind of discussing the implications of it, erm, things like that. So probably kind of like mainstream media is probably how I heard about it for the first time."
  • Interviewer: "And did you, did you see it as something that, that would impact you personally?"
  • Respondent: "Erm, to be honest, no, I didn't really see it as like, I didn't see it as kind of something that would be available to the public, if that makes sense. I kind of, I kind of believed that it would be kind of more owned by like corporations, by like people who were involved with like tech and like computers but nothing really involved in with me per se. Like I didn't think it would be accessible to the general public, let alone myself so."
  • Interviewer: "Right, okay. What did you feel about it? What were your first sort of thoughts around it when you heard about it?"
  • Respondent: "Erm, I think my first kind of initial thought was kind of more like fear, like an anxiety around it, erm, especially like with, like I said previously, the pop culture and like mainstream media. Like it's usually depict as kind of a bad thing, like AI kind of evolving to be kind of better and superior to humans and taking over and things like that. So initially I kind of had a bit of a fear response regarding like what the future of it would be, kind of how it would be integrated into like corporations and like day to day life and stuff like that. Erm, so it was kind of like an initial like fear response regarding it."
  • Interviewer: "Okay. And thinking about that, could you, erm, could you focus that? In terms of that sort of anxiety, if I asked you what that was about, would you be able to sort of put your finger on what that was?"
  • Respondent: "Yeah. So I think it's kind of like the anxiety of like the uncertainty of it, kind of like, like I said a minute ago, like if it's like the future of it and it kind of taking over, like it's a lot of uncertainty about like where I'd fit in that picture, like where humanity as a whole fits in that picture, kind of the future, if like I was to use it, like how would I use it? Like is there certain like lines that can be crossed when using it? Like what's the difference between us and AI if it gets to a certain point that it can mimic humans like really well, like where does humanity fit in that picture? Can it just take over from us? Like are we, is our existence and everything to do with that like relevant anymore if there's something out there that can replicate everything that we do, er, like our interactions with each other, like the way we write, the way we speak. Erm, so it was kind of more an anxiety and a fear of the uncertainty and where our place is as humanity and that uncertainty regarding the future of it."
  • Interviewer: "That's really interesting that you're seeing that in the big picture, erm, and I wondered if you sort of projected forward, how you saw that potentially impacting your own life and your own sort of future."
  • Respondent: "Erm, I'd say kind of like, I mean there has been talks about this in kind of like media, like I said previously, but kind of like there is a possibility of AI kind of taking jobs, kind of taking like, er, things like that. So I think I'd be more concerned regarding like me, myself, going, obviously being an undergraduate, like I'm kind of aiming towards a future career, erm, and whether like AI, if it can replicate basically what I could do, like would I even have a job to even go into if like AI is kind of taking those positions, cos obviously it's, it would be cheaper to have like an AI kind of robot, like less maintenance compared to like a human worker. So I'd be more concerned about my future in terms of my career, in terms of kind of like what I'd be able to give to society, if there's something else that can mimic what I can do then where do I fit in that."
  • Interviewer: "Absolutely. Again, what I hear is, is an anxiety both for you as an individual but also how, erm, how society and the structure of society could alter."
  • Respondent: "Yeah."
  • Interviewer: "And of course without really understanding what that might look like."
  • Respondent: "Yeah."
  • Interviewer: "So that's, that's kind of the basis of the anxiety?"
  • Respondent: "Yeah, the basis of it is kind of the uncertainty of where everything fits."
  • Interviewer: "Yeah. No, that's really understandable. So in terms of your first interaction with AI, erm, could you tell me about that?"
  • Respondent: "Erm, I'm trying to put my finger on it. I think the first time kind of AI was introduced to me was I think like on social media and kind of like you see on Instagram like a lot of like, erm, I think they call them deep fakes and like fake pictures as kind of like I've seen these images on Instagram say of like a tragedy or say of like a celebrity and it looks like a real picture but when you kind of go through the comments or like you go through like the sources of where the image came from, and it's actually an AI generated image. So that's kind of the first time I was actually like introduced to it being like used like there and then. Like I said previously, in the mainstream media when they talk about it, like it's kind of like a futuristic thing, it won't actually exist until like 10-20 years in the future. And then one day I'm just on social media like looking through and I see like oh, that image isn't it real. Like that's been created by a computer that's been created by AI. Erm, so that's kind of like the first time I've really been introduced to it, is kind of on social media, like Instagram, Snapchat. Erm, notable at the moment, Tik Tok have been quite like, I've seen a lot of it on Tik Tok. Erm, AI has been kind of mimicking like voices, been mimicking like celebrity voices and doing like songs and things like that with and it's, yeah, that's kind of like the first time I was kind of introduced to it, is through social media."
  • Interviewer: "And how did that, how did that affect you?"
  • Respondent: "Erm, I was kind of like taken aback when I first kind of saw it. It was, erm, it was quite a shocking realisation cos I didn't realise that it was so, I didn't realise it had kind of advanced as much as it had. Like I didn't realise that it was actually there, like it, you know, for the general public to use. Like I said earlier, I didn't think it would be something that the general public would have access to. And yet when I first saw it on social media, I was really kind of shocked and taken aback cos I was like, 'Ah, this is something that we all can use now.' Like I didn't realise that they had made such an advancement on it, to the point that it is just very much present, very much there on my like social media feed. Erm, and it kind of felt like, yeah, it was a moment of like shock and realisation as to like how far we'd come with it and with that, that it is present."
  • Interviewer: "Did it change anything about, cos where you started was this sort of, erm, your initial, when you initially sort of heard about AI, it felt like it was something distant if you like, something, erm, something that potentially countries or big corporations might, might be utilising but something that may be a little remote from your every day life. And then, and then when you've got this sort of direct exposure through social media, then it becomes very much personal. It's very much, er, in your space. I just wondered if that had changed, changed your, your view on it or your feelings about things when you first, your first direct encounters with it."
  • Respondent: "Yeah, I think it definitely kind of, like you said, like it definitely changed the way I saw it because like you said, I honestly thought that it was something like that countries or corporations would have access to . So the fact that it was in my personal space kind of, it gave me a sense of more control than I originally… so I have a fear of like uncertainty about things. And I think it being accessible to me and being on my like social media feed, it felt like I had access to it so I kind of had a bit more control about it. Like I could use it if I wanted to. Like I can use it in my own ways. I could use it like to make posts and things like that. So it definitely changed my like opinion on it, being this big kind of scary like thing that only like the big people could like have access to, like the corporations, like things like that. And it kind of made me feel like, 'Oh, you know what? I actually have like control of this. I can use this like freely if I wanted to.' So it changed like the perception of it being like this far away that I couldn't reach, to being like something I could use in my own day to day life if I wanted to."
  • Interviewer: "Did it change anything or how did it feel that you were seeing things on social media and that that appeared real but weren't if you like?"
  • Respondent: "Yeah. It kind of, I mean I think when you use social media anyway, like you do have this sense of like nothing's real. Like you know a lot of things can be photo shopped. You know, like people kind of post the best parts of their life. Like you don't see a lot of kind of what is real anymore, especially like on social media. But I think seeing AI being used to blatantly on social media quite freely without any like… like a lot of people when they post as well, they don't even say like this is created by AI. You kind of have to dig to find out. So I think seeing that like on social media, kind of solidified the idea but like most things we see, especially things online, isn't necessarily based on reality. And it kind of, there was this kind of like separation kind of in terms of like a detachment from like my personal feelings towards like most things I see online now. Like I think the introduction of like AI on social media kind of distanced my idea from like anything ever being real. Like now when I use social media, especially because of AI, when I use social media, I don't see anything on there. I take everything with a pinch of salt. Like to me, if I saw an image now, I would very much question if it was real. And there's quite of a dissonance between me and that. Like I don't care if it's real anymore, if that makes sense. Like if I wanted to see the real world, I'd just go out there. Whereas now when I use anything online, it's not, I don't really care for it anymore because like I said, like there is no semblance of anything being real anymore, especially with the like introduction of AI."
  • Interviewer: "That's really interesting and I… is that something that's, is that something that you're conscious of developing over time, that, I'm loathe to call it a cynicism but, but kind of a, I think you called it detachment."
  • Respondent: "Yeah."
  • Interviewer: "I just wondered how that had developed for you."
  • Respondent: "Erm, I'd say like that quote, that developed for me using social media from quite a young age. Like you do see a lot of like… it was introduced like years ago, kind of the idea of photoshop. Like a lot of celebrity images aren't real. Like a lot of like for example the Kardashians and people like that, like they would use photoshop on like a lot of their images when you see it on tabloids and magazines. So I think that kind of, from a young age, using social media, like it's kind of like, it was solidified anyway. But I think from the start, it was solidified to me as like only celebrities use this and like only like the people who own the magazines, they use photoshop. They use like the AI to make their images look good. But I think as it's developed over the years and like with more like AI appearing quite frequently in social media and like things like that. It's kind of like changed from my idea of it just being celebrities and just being like these high up people to it being actually now everybody is probably like changing their images. A lot of people are probably like editing their images. A lot of people are like posting fake things, whether that's with the use of AI or any other like, erm, editing like software like photoshop and things like that, which, coincidentally, also uses AI now so practically all of that is AI. Erm, so it kind of developed from being just only certain things are fake, only like the celebrities post fake things, to it kind of developing towards like everything you see is kind of fake and that detachment grew a lot more stronger over time then like I said, with the introduction of AI and seeing AI on social media. I think that was kind of like the final straw of kind of like truly being like nothing on here could be real and it doesn't really matter."
  • Interviewer: "Right. That is really interesting how, how you… cos it's almost as if, so for you, AI is almost a progression on from things like photoshopping where you're, you're enhancing or changing the appearance of something to create a certain impression. And so from that extent, that's, you saw this as an extension of that, that progress?"
  • Respondent: "Yeah."
  • Interviewer: "Okay. Erm, have you used AI yourself?"
  • Respondent: "Erm, I've used, I've used it… I'm trying to think. I have used it in some senses. So I've used it probably the most on things like Tik Tok and Snapchat cos, erm, Snapchat now has this, erm, feature, they only brought it in probably like a couple of months ago. It's called like My AI. So it's essentially, erm, like an AI tool and you kind of just text it as if you were texting an actual person. The responses aren't as good as what you would probably get from a normal conversation with somebody but you can ask it questions, you can ask it like opinions on things, you can ask it to like write you a story if you wanted to, so that's probably the first kind of instance where I started using it. I think I saw it come up on Snapchat and I kind of wanted to test it, like I wanted to see how real and how kind of integrated it all was. So I think that was kind of the first time I've used it. And then in conjunction with that, like Alexa cos that's kind of like an AI tool. So when you speak to Alexa, you can kind of tell it to play music, you can tell it to like Google things. So that's probably again like Snapchat in conjunction with Amazon cos, for me, it was around the same time. Erm, that was probably like the first instance I had of using AI. And I think when I used it for the first time as well, I didn't necessarily use it out of need. It was kind of more out of curiosity. So I think I really wanted to kind of see, like I said a minute ago, kind of like what the limits of it were. Like I wanted to see how much it could do and how like real it was in terms of like being able to do tasks for me, like Google things for me, like how real was it in like having a conversation, how like much of a human like kind of thing it had to it."
  • Interviewer: "Okay. And how did you find that?"
  • Respondent: "I found it surprising. I didn't think, erm, I didn't think it had like advanced as far as it had. Like I kind of was expecting to kind of like, you know, throw a few random phrases at it that didn't make sense and it would kind of like not understand and kind of like shut down. But in reality it actually has, it's progressed a lot more than I originally thought. So I was kind of shocked when I started interacting with it, to find that it has actually progressed quite a lot further than I originally thought. Erm, I think I was taken aback by that and I was kind of a bit, it was a bit, like in essence, a kind of fear kind of took over me at that point because I realise now like I have this Amazon in my house. Like it's in my house. It's with me every day. The same with my phone, like with Snapchat, with like things like that. I was like oh well, it's kind of too late now, it's kind of in my life now. Like if, for some reason, it advanced to a certain point, like it's already in the door, it's already in the house. Like there's really nothing I can do now if anything goes wrong with it. So there was kind of an element of shock and fear I'd say for like the first time using it."
  • Interviewer: "So did you feel that it's kind of an inevitable process?"
  • Respondent: "Yeah. I think it was, now that I've realised like how quick we've moved with it, I think it was like inevitably gonna be in everybody's households. Like it's gonna be everywhere by like, I'd say, I mean if it isn't already, it probably is everywhere already, if not by the end of the year, it would be in most people's houses, it would be in most like supermarkets. I mean you see it on like the checkouts. Like that's a form of AI so like, you know, it will be like everywhere. And I think my opinion of it being like oh just for the big corporations and just for this kind of transitioned into oh no, I can use it, and then transitioned further into oh wait, it is like everywhere and everybody can use it, which is kind of scary but."
  • Interviewer: "That's a really interesting sort of process of, of working through that in terms of its interaction or contact with your own life. Erm, so I guess, erm, your, so you've really used it, erm, in a, I guess, a social capacity?"
  • Respondent: "Yeah."
  • Interviewer: "Sort of fairly low level kind of where you're, curiosity is the driver rather than anything."
  • Respondent: "Yeah. Like I'm not really using it for any kind of like big kind of things. I'm not really using it to write essays. I don't really use it in any of those ways. It was mainly just kind of a curiosity, see what it can do for me, see how far it's progressed and that was kind of the baseline level of it."
  • Interviewer: "So you've mentioned, erm, you've mentioned, so it sounds like the majority of your AI use has been through your phone, would that be fair?"
  • Respondent: "Yeah. Yeah, I'd say that's fair. My phone and probably my laptop, erm, cos there is quite a few like AI, erm, software like Grammarly and things like that."
  • Interviewer: "Oh, I was gonna say have you used it, you know, on the laptop as well cos it sounded like it was probably the phone but."
  • Respondent: "Yeah. It is mainly my phone to be honest. It is like kind of things like, like I said, like Snapchat and Instagram with the occasional use of it being on my laptop for any applications like that."
  • Interviewer: "Right, and is that something, in terms of the laptop then, is that something that you sought out?"
  • Respondent: "Erm, what do you mean by that?"
  • Interviewer: "Er, did you, did you sort of think how can AI help me with whatever, grammar, you know, things like that?"
  • Respondent: "Yeah, well I kind of, the way I started using it on my laptop is kind of more through like word of mouth. So I kind of had a few friends at uni kind of being like, 'Oh by the way, you should probably check out this application. Like it's good for like your spell checks, it's good for your grammar if you're ever writing an essay. It can kind of like do a check over for you kind of just before you submit it.' Erm, so I thought, 'Oh, that sounds interesting.' So it kind of moved from like a curiosity if being like, 'Oh, I can actually use this functionally.' Erm, so I started kind of like looking at it on my laptop, like looking at kind of what, er, applications would do that purpose and like Grammarly kind of being the main one that I kind of, is mainly used and has kind of like got the most use for me as well. So I kind of started using Grammarly. Erm, I don't really use it that often. It's kind of more like every now and then if I've got an essay that I need to submit and I kind of want to just do a final check over to see like if your grammars all right, if like your punctuality is all right, like I'd probably just pop it in, go and scan it. It would be like oh, you've got a few spelling errors. I'd just do that and then I'd turn it in to like uni. So it kind of serves as more like a, in my life it's only like a, the usage is quite low in terms of like it just does like little bits and bobs here for me. But I do know some people at uni who have used it quite extensively to kind of like direct them in terms of essays, direct them how to like have an essay flow. Erm, I've heard of some cases, like they're not friends of mine but I've heard of some people who have just asked an AI service to write a whole essay for them with like, you know, certain contexts added. And like they've just turned in an essay that isn't theirs but is written by an AI, erm, which is quite interesting to think about in terms of like the university and education context, like where do we go from here in terms of that. But for my use personally, it's only been like small base line things like that."
  • Interviewer: "What were your feelings on that?"
  • Respondent: "Erm, what in terms of my own use or in terms of other people's uses?"
  • Interviewer: "Well in terms of that example that you just gave me where somebody's turned in an essay written by AI. How does that, how does that sit with you? How do you feel about that?"
  • Respondent: "It's kind of, I have this kind of, erm, I don't know how to say it, like it kind of feels like this was kind of the progression that it was leading to with seeing AI, how it's progressed over the years and seeing that it can do things like this, like check your essays, like write your essays for you. It was kind of like in my head, as soon as I saw that I was like, 'Yeah, somebody's defo gonna use that, erm, to write their work.' And it was kind of more like in my head when I saw it, I was like oh, like it was an inevitability. Like I wasn't too shocked by that at this point cos obviously like I've had the year of like seeing it progress. Like and at that point when I got to uni, I was like, 'Right, this is kind of like how it was going to be.' I think more what my concern is now is kind of the future of like the education system in terms of like how, like coursework is gonna be done, how essays are gonna be done because if you can't discriminate between what somebody's written and what an AI has written, how are universities gonna like mark that. And I've heard of some cases, because it's quite funny as well, is that the fact that the uni uses AI to actually mark students work. So you have like Turnitin, that kind of like, you know, analyses somebody's work to see where, if there's any like replication from somewhere else, if there's any like, erm, like cheating and things like that. So I've seen a few cases where people have handed stuff in to Turnitin and Turnitin, being an AI, has mistaken somebody's work for AI. So now we're kind of at a standstill where we don't know who's written what and what's been written by an AI. Erm, so I think now my feelings on it are kind of more of a fear but not for myself, not for my friends cos we're almost finished with uni but more for the future of the education system, the future of coursework, the future of like how exams are gonna be done, if they'll have to be reverted back to in person or if we can continue going forward with things being done on computers and things like that with the advancement of AI."
  • Interviewer: "Yeah, I can see that. So what's quite interesting is that you've, you've spoken about, erm, where you sort of started to use AI and you spoke about being surprised, in some respects, at how developed it was. And I wondered if it was, what had held you back potentially from using it more extensively in, in your work. Erm, and I don't necessarily mean in a, in a way where, you know, where for example, you know, getting an essay written for you, that might. But there are many things it could have potentially done that, that, where you've spoken about it, erm, you've spoken about it as a tool before and you could see that, erm, that natural progression with something like, like photoshop to image enhancement with AI. And I just wondered if there were, erm, maybe barriers or things that were holding you back from using some of it, what that might be."
  • Respondent: "Yeah. Erm, I've been asked this actually by a lot of my friends, erm, cos they use it quite extensively so they say like, 'You know it can do so many things for you, why do you not use it?' I think for me it's more like it's kind of probably a moot point for me to do this but I kind of want to prove my own efficiency. Like I want to prove that I can actually do something. Like I was saying earlier, like my fear regarding AI was that it would eventually take over, as in replace humanity. So I think my lack of use of it is kind of coming from my own self of being like I want to prove that I can actually do something, like I can do something as well as an AI. I can probably not but I would like to try and like prove that anyway and to kind of like… using AI I think is kind of like, in my own head it's kind of like a cheat. It's like I'm not actually producing my own work, I'm not actually kind of doing my own stuff, even if it is like putting on music in the house. Like I can easily go over to my phone and do that. I don't need an AI to do it. Like it comes from kind of a place of proving that I can still do things without the use of computers and AIs which, like I say, is probably a moot point because we're heading towards a space where we probably won't have to do anything anymore. But yeah, I do have a lot of friends who use it and they're like, 'Why do you never use it?' and for me, it's just I just don't want to. Like I want to be able to do things myself without the help of a computer."
  • Interviewer: "Does it feel, er, as if the, and you've kind of spoken about this inevitability."
  • Respondent: "Mmm."
  • Interviewer: "Erm, I get the sense that you, you've spoken about things that potentially, erm, will get lost."
  • Respondent: "Yeah."
  • Interviewer: "That's how it feels. Is that, would that be right? Is that, is that what you're talking about?"
  • Respondent: "Yeah."
  • Interviewer: "That somehow you see things potentially getting lost as a result of it?"
  • Respondent: "Yeah, definitely. Like human non-creation, like art, music and stuff like that, even AI is doing it. Like you kind of lose like a lot of like the humanity and everything."
  • Interviewer: "Okay. So it's, so those are the areas that, that you're concerned are going to be potentially adversely affected."
  • Respondent: "Yeah."
  • Interviewer: "Could you tell me about that?"
  • Respondent: "Erm, yeah, I think it's kind of, like I've seen like, like I said, on like social media and stuff, like a lot of AI can kind of mimic art now. It can kind of like, if you give it like a picture or if you kind of say oh, draw me a picture with like this context or like this and this, it will just do it. And the same with music, it can just create music now and stuff like that. So I think particularly in the area of arts and things like that, like it is kind of a shame cos like you do lose out on that individual like human creation. Like you'd see like art pieces from years ago and you'd be like, 'Oh, like this is like insane. This is amazing,' but now like we've got computers that can do things practically the same. And it's, kind of like we were saying, like it's a loss of like that like humanity and everything. It's the loss of like individual creation. And like everything we see now, it could be AI, it could be human. Like we won't be able to tell anymore which I think is kind of sad because you'd see like a really amazing piece of art work of like 100 years ago and it would be like, 'Wow, that is kind of like human creation at its like pinnacle,' but now you go into an art gallery, you see something and there's no way of telling if that's human or not which is kind of a sad loss really but yeah."
  • Interviewer: "That's really interesting. Sort of the last part really, erm, is reflecting backwards from where you are now cos you've been through that, you know, you've been through at a really pivotal time, the, erm, the explosion of AI really through a point where there was none. Erm, how do you see the differences in life before AI?"
  • Respondent: "Erm, it's kind of hard to say for me I think because I was growing up at such a time where like you have like your supermarket checkouts that out like done by a human. Like I think throughout my life, it's kind of been more like of an integration that's kind of like, I mean like you said, it has exploded in the last like couple of years, especially like in mainstream use. Erm, I feel like, I feel like some things have changed. I feel like a lot of things kind of in terms of like, like I was saying earlier, people writing essays, like people actually doing their own work, erm, people like… for example my partner, he does use a lot of AI in his stuff, like as in he can use it to like just things for him, it can get the references up. So he can plug a study into an AI thing and it will give him the reference and it will like write it out in like an APA format if he needed it to. So there are like a lot of people who kind of use it as shortcuts. But in terms of like day to day life at the moment, for the time being, because it's not like as advanced as to say like we've got like robots walking around with us. It's not, it's not changed too much I don't think. Like I feel like there is things that have changed definitely in terms of like, like what we see online and what we see like created by AI. But in terms of like day to day life, if you're not on like social media very frequently, if you're not using like your computer and technology as frequently as like some people, like I wouldn't say there's been massively that much of a shift. Erm, it's kind of gradually come about over the last like 20 something years that I've been alive. Erm, I imagine if somebody's born like, like 30-40 years ago, there'd be a lot more of a change that they would have seen. But for me cos I've kind of grown up within that age of technology, within like the age of like seeing like some things integrated, it's not been as much as like a before and after. It's kind of just been like a slow gradual thing that's just like yeah, some things have changed but like nothing massively, like nothing massively compared to the other major events that have happened, like the pandemic and stuff like that. Like if you were to compare it to things like that, then there's not been that much of a change before and after."
  • Interviewer: "Right. Okay. Well that's, I think that's everything that we need to cover. Erm, so, [Name], thank you very much for taking part, er, it's been really, really interesting. I really appreciate it. Erm, just before we finish, I'd like to remind you that what will happen with the transcript, we'll anonymise anything in there that was personalised to you. Erm, and it won't have any impact on any of your university work. So just as a reassurance. I'm gonna stop the recording. Erm, there we go."