Unfiltered:Justificativa e Relevância

Unfiltered: Banco de dados

• 2019 Brazilian National Health Survey (PNS);

• target population: self-reported as smokers in PNS 2019;

• number of cigarettes smoked per day; education level; race/skin colour; private health plan coverage.

Unfiltered: Testes de Significância:

• private health plan coverage: Wilcoxon–Mann–Whitney test.

H0: the distributions of both populations are identical; H1: is that the distributions are not identical.

• educational attainment and race/skin color: KruskalWallis test;

H0: medians of all groups are equal; H1: at least one population median of one group is different from the population median of at least one other group.

Unfiltered: number of cigarettes smoked per day

Unfiltered: Private health plan coverage:

Unfiltered: Private health plan coverage:

    Design-based KruskalWallis test

data:  P05402 ~ I00102
t = -0.94228, df = 3935, p-value = 0.3461
alternative hypothesis: true difference in mean rank score is not equal to 0
sample estimates:
difference in mean rank score 
                   -0.0137362 

Unfiltered: educational attainment:

Unfiltered: educational attainment:

    Design-based KruskalWallis test

data:  P05402 ~ VDD004A
df = 6, Chisq = 5.9261, p-value = 0.4317

Unfiltered: race/skin colour:

Unfiltered: race/skin colour:

    Design-based KruskalWallis test

data:  P05402 ~ C009
df = 4, Chisq = 25.67, p-value = 3.822e-05

  C009   P05402        se
1    1 14.13265 0.3169038
2    2 12.44021 0.4547649
3    3 15.24367 2.6940833
4    4 12.43599 0.3042334
5    5 11.14804 2.0917019

Unfiltered: Conclusões, Limitações e futuro do artigo:

• “Contrasts with the expectation of a clearly decreasing social gradient among more educated smokers.”

• It is plausible that socially vulnerable groups concentrate a higher prevalence of smokers, but that among those who remain smoking, daily intensity does not follow the same gradient, reflecting combinations of differential cessation, long-term consumption patterns, and types of products used.

• Mais Análises com mais variáveis e comparar quantidade de fumantes x intensidade do hábito de fumar.

Threat or Tool: Objetivo e Hipótese

Threat or Tool: Justificativa e Relevância

Threat or Tool: Banco de Dados:

• Secondary data from the public AnthropicInterviewer corpus, originally collected by Anthropic with informed consent for transcript release and research use;

• What 1,250 professionals told us about working with AI: MAS ENTREVISTADOS POR UMA IA

• The macrodata includes full transcriptions of 1.250 professionals, split into three major groups: general workforce (1.000 participants), scientists (125 participants) and creatives (125 participants.

• For this study’s goals was delimited to the creative workers group (N = 125). This subsample includes stories from writers, musicians, composers and visual artists, in order to extract responses regarding ethical questions and the psychosocial impacts of AI on daily creative industry work.

Threat or Tool: Metodologia:

• Reflexive Thematic Analysis (RTA) is a qualitative analytical method developed by Virginia Braun and Victoria Clarke;

• Researcher’s subjectivity is classified as an essential resource and main knowledge production instrument;

• The methodological outline proposed by this study applies human intersubjectivity onto the interpretation of large-scale generated data.

• The following topics were extracted: i) AI as creative work assistance; ii) Author- ship and authenticity as risk; iii) Human control over AI; iv) Limited trust on AI answers;v) Fear of human substitution and professional devaluation.

Threat or Tool: Resultados i) AI as creative work assistance

“I mostly like to get unique details rather than fully fleshed-out concepts.Just something to get the ideas jumpstarted for me. It felt great to have ideas to pick from instead of sitting around trying to do the grunt work myself.” (creativity 0060)

For the positive side, in the early stages of design or writing, AI is seen as an invaluable tool for breaking through initial creative blocks.

On the negative, true artistic breakthrough often happens during the slow,frustrating process of working through a creative block. By using automation to escape this friction too quickly, creatives may miss out on deep learning opportunities and the accidental discoveries that define organic artistic evolution

Threat or Tool: Resultados ii) Authorship and authenticity at risk

“I don’t use generative AI in my work, because at that point, what is the point of me as a creative? And what is the work without the artist? Simply entering in a prompt to generate an image that is ultimately simply an average of images scraped off the web is not creative. The end result lacks substance.” (creativity 0036)

For optimistic viewpoints, authenticity is kept by human curation and editing. The tool suggests, but the creator refines and validates.

For negative viewpoints, AI threatens the integrity of art, as it separates the aesthetic result of effort and the life experience needed to produce creative work. The outsourcing of the creative core is seen as a moral violation.

Threat or Tool: Resultados iii) Human control over AI

“Rarely have I found the AI particularly to be ’creative’ if you will, specif- ically as I am not sure I could say that the system prompt’s I have written allow it to exist outside of typical artistic norms. I would say I have al- most exclusively been directing it, and found that to be the number one pain point in terms of my direction being lost from say prompt to image generation or image editing.” (creativity 0014)

Creatives who hold positive views towards AI usage ensure they remain the driving force, many creatives establish strict self-imposed rules. They limit AI participation to early conceptual stages or mechanical tasks, drawing a clear line that prevents the machine from making core creative, rhythmic, or melodic choices.

However, negatively, there is a subtle danger that over time, in which fatigued creators will compromise their original standards. Instead of pushing through multiple iterations to achieve their precise vision, they may settle for a generic, automated option, quietly surrendering their creative intent to the dataset limits of the software.

Threat or Tool: Resultados iv) Limited trust on AI answers

“Right now, I don’t fully trust it to edit a work to completion, and its design work often requires a fair amount of oversight and re-tooling to get right. Hopefully in a few years its refinements will make it an even more useful tool.“(creativity 0015)

The positive outlook is that rather than accepting outputs blindly, copywriters and researchers build multi-layered review processes to screen for hallucinations, evaluate tone accuracy, and check stylistic consistency.

On the negative side, the lack of trust also highlights a significant operational drawback: AI can be incredibly inefficient for high-stakes projects.

Threat or Tool: Resultados v) Fear of human substitution and professional devaluation

My concerns are with people and companies trying to abuse AI by using it as a replacement for artists and creatives like myself who’ve spent years training and honing our skills.” (creativity 0009)

From an optimistic viewpoint, AI tools lower the steep cost and technical barriers that historically kept independent creators out of advanced industries.

However, the most alarming structural risk is the destruction of junior-level career paths. By automating baseline tasks—like early copywriting drafts, fundamental editing, or simple graphic layout work—the industry cuts off its traditional training pathways.

Threat or Tool: Discussão e Conclusão

• While optimistic views tend to frame AI as a support, acceleration and magnification tool, pessimistic views associate it with authorship loss, authenticity and trust weakening and professional devaluation.

• The debate about generative AI within creative work mustn’t be reduced to the opposition between innovation and resistance, with a dispute about the meaning of creativity within the context of automation being at play.

• I can be perceived as a supporting tool, but such support is faced with questions related to authorship, trust, control, copyright and recognition.

• The Reflexive Thematic Analysis allowed for visibility to be raised towards such tensions by interpreting data as situated narratives, in which creative professionals continuously negotiate the ethical limits of the technology within daily work.