One PhD, five papers, five posters
A publication pipeline from the thesis, with a poster for each paper in its own visual style. They are honest about where the work is: these are work-in-progress posters showing questions, data and design, not findings that don’t exist yet.
What could be published, and when
The order is set by when the material is ready. Paper 1 can be drafted now from Chapter 3. Journal names are suggestions to discuss with supervisors.
The pure relationship, minus reciprocity: AI companions and neuroliberal wellbeing
Builds the loop argument from Ch3: Giddens’ pure relationship, artificial sociality, therapeutic culture and Isin’s neurotic citizen, set against relational wellbeing (White) and Javanese ethics.
Body politic, body personal, body delegated? The Guardian and AI chatbot wellbeing, 2023–2026
Continues Sointu’s (2005) trajectory in the same newspaper: what counts as wellbeing, who lacks it and what the remedy is, across emergence, normalisation and containment.
Parenting the chatbot: family journalism and the relocation of responsibility for wellbeing
How “Family & Tech” moves AI chatbot wellbeing from regulation into the home, and which parent the column addresses, building on Sointu’s finding that women were made responsible for wellbeing at home.
Sungkan and the machine: how Indonesian producers of digital mental health define wellbeing and governance
Founders, psychologists and HIMPSI on where self-care ends and clinical care begins, and what governing AI chatbots would require where no framework exists yet.
Reading responsibility in the grammar of news
A short methods piece showing the four-level van Dijk protocol on AI news: macropropositions, schema slots, nominalisation and source types as evidence of who is held responsible.
Plus: one output for Indonesia
A Bahasa Indonesia policy brief for HIMPSI and platforms, and a public piece (e.g. for The Conversation Indonesia). This meets the LPDP expectation of contributing back, and it is something to point to in job applications.
Five styles, A-series proportions
Click a poster to enlarge it. “Print” gives an A4 proof; scale to A0/A1 in the print dialog. Posters always print on light paper, whatever the site theme.
The Wellbeing Gazette
Who is made responsible when we hand our wellbeing to a machine?
News discourse on AI chatbots, 2023–2026, and the Indonesian case
The scale
ChatGPT users a week who show explicit indicators of possible suicidal planning (OpenAI, Oct 2025: ~0.15% of ~800M).
The gap
Weeks-long NHS waits in the UK. In Indonesia: stigma, private psychiatric insurance, and pasung.
The argument
AI companions copy the form of the pure relationship but remove its reciprocity. Users look for wellbeing in the same relationship that keeps producing the need for it. News decides who must manage that loop, and it does so before regulation arrives.
Three strands
RQ1 The Guardian, 86 articles identified so far.
RQ2 WSJ “Family & Tech”, 20 columns.
RQ3 Riliv founders and psychologists, and HIMPSI, in Indonesia.
Method
van Dijk news discourse analysis · reflexive thematic analysis
Body politic, body personal, body delegated?
How The Guardian has redefined wellbeing around AI chatbots, 2023–2026
Corpus
Guardian articles found for “AI chatbot and wellbeing” so far, 1 Jan 2023 – 31 Dec 2026, before screening.
Stages
Emergence → Normalisation → Containment, adapted from Downs’ issue-attention cycle, with boundaries drawn from the corpus.
We code
What counts as wellbeing · who lacks it · the remedy · who acts · who is quoted · what goes unsaid
Why it matters
The definition that wins in public will shape what regulation counts as harm, and who it holds responsible.
Parenting the chatbot
How family journalism moves AI chatbot wellbeing from public regulation into the home
Julie Jargon, “Family & Tech”, The Wall Street Journal
Each dot is a column (Jun 2023 – Oct 2026). Colour shows theme: harm, clinical, children, everyday life, safeguards.
columns in the corpus so far, one consistent writer
of them published since October 2025: attention arrives with containment
Questions
- Who is the column’s subject: the child, the parent or the company?
- Which parent is addressed? Sointu found news made women responsible for wellbeing at home.
- Do parental controls turn supervision into the remedy?
Sungkan & the machineSungkan dan mesin
How people who build and license digital mental health in Indonesia understand wellbeing, responsibility and governance
Who speaks
Riliv co-founders · Riliv psychologists (focus group, then individual interviews) · HIMPSI, the body that licenses psychologists under UU 23/2022.
Hypothesis to test
A chatbot may appeal not because it meets an individual need but because it solves a relational one: it lets a person speak without disturbing harmony. It may also extend silent suffering.
The pure relationship, minus reciprocity.
This paper brings these literatures together to argue that AI companions are where neuroliberal wellbeing is most complete: the relationship is perfectly satisfying, never mutual, and designed by a company.
From a 15,000-word chapter to a poster read in 90 seconds
Choose one claim
Write the sentence you want someone to remember from across the room, e.g. “News places responsibility for chatbot wellbeing in the home before regulation arrives.”
Choose one figure
Pick from the figure options. Use the arc for RQ1, the dot timeline for RQ2, the four norms for RQ3 and the loop for theory.
Cut to 300 words
Three boxes of about 60 words (question, data, what we look for) plus the title. Anything longer belongs in the paper.
Lay it out in a Z
Title top left, figure in the middle, takeaway bottom right. One palette: teal, gold, terracotta.
Test at 25%
Print an A4 proof (the buttons above) and read it at arm’s length. If the title and figure don’t carry the claim, simplify.
Prepare the 60-second pitch
Problem (1M+ a week) → question (who is made responsible?) → design (three strands) → why Indonesia. End with an invitation to talk.
Pick the venue
UK: MeCCSA and the Goldsmiths MCCS research events. International: ICA, IAMCR, ECREA, AoIR. Asia: regional communication association meetings.
Reuse it
The same figure and claim become the paper abstract, a conference slide, a LinkedIn or Bluesky post and a line on the CV.