Dr Zuleyka Zevallos
@othersociology.bsky.social
Applied sociologist on Wurundjeri land. I write about social justice: https://othersociologist.com/
She/ her
She/ her
Hi Ally. Delicious Monsters - Liselle Sambury
October 13, 2025 at 4:12 PM
Hi Ally. Delicious Monsters - Liselle Sambury
My analysis shows how AI discourse overstates the technological utility of AI products. This discourse seeks to eliminate competition by exploiting and undermining the arts and humanities, and hiding the true costs of automation.
/End thread
othersociologist.com/2025/09/03/a...
#ai #sociology
/End thread
othersociologist.com/2025/09/03/a...
#ai #sociology
Artificial Intelligence at Work
Despite dire predictions, researchers predict that few industries will lose a substantial number of jobs due to artificial intelligence (AI). Instead, AI is more likely to alter the way humans work…
othersociologist.com
September 3, 2025 at 3:13 AM
My analysis shows how AI discourse overstates the technological utility of AI products. This discourse seeks to eliminate competition by exploiting and undermining the arts and humanities, and hiding the true costs of automation.
/End thread
othersociologist.com/2025/09/03/a...
#ai #sociology
/End thread
othersociologist.com/2025/09/03/a...
#ai #sociology
A recent Microsoft study claims that AI will take over 40 professions, including historians and translators. I show that the study fails to control for demographic patterns. The study focuses on a narrow list of functions and outcome measures that raise issues of validity and ethics
#ai #sociology
#ai #sociology
September 3, 2025 at 3:08 AM
A recent Microsoft study claims that AI will take over 40 professions, including historians and translators. I show that the study fails to control for demographic patterns. The study focuses on a narrow list of functions and outcome measures that raise issues of validity and ethics
#ai #sociology
#ai #sociology
JSA does not account for empirical studies showing the high rate of error in AI models. I show how the JSA report is biased, particularly by overestimating the likelihood of AI success, when a recent MIT study shows 95% of AI companies fail to generate a return on investment
#ai #sociology
#ai #sociology
September 3, 2025 at 3:02 AM
JSA does not account for empirical studies showing the high rate of error in AI models. I show how the JSA report is biased, particularly by overestimating the likelihood of AI success, when a recent MIT study shows 95% of AI companies fail to generate a return on investment
#ai #sociology
#ai #sociology
The JSA study relies on a theoretical model, which asks ChatGPT-4 to rate how well it can perform a list of job TASKS. This model has not been tested empirically. JSA applies this AI model to an outdated list of occupations. JSA used AI to conduct its analysis and write the report.
#ai #sociology
#ai #sociology
September 3, 2025 at 2:57 AM
The JSA study relies on a theoretical model, which asks ChatGPT-4 to rate how well it can perform a list of job TASKS. This model has not been tested empirically. JSA applies this AI model to an outdated list of occupations. JSA used AI to conduct its analysis and write the report.
#ai #sociology
#ai #sociology