The Korea Advanced Institute of Science and Technology (KAIST) announced on the 28th that a research..

The Korea Advanced Institute of Science and Technology (KAIST) announced on the 28th that a analysis staff led by Professor Choi Moon-jung of the Graduate School of Science and Technology Policy quantitatively analyzed that delicate stereotypes about the aged are inherent in the sentences generated by OpenAI’s ChatGPT-4o, a generative synthetic intelligence mannequin.

The analysis staff collected 900 texts generated by GPT-4o utilizing a impartial immediate to explain the traits of the 10-year age group from the age of 10 to 90.

Afterwards, the stereotype content material mannequin (SCM, a principle that explains perceptions of folks or teams in two dimensions of ‘warmness’ and ‘competence’), a consultant principle in the subject of social psychology, was utilized and analyzed.

As a outcome, it was confirmed that the aged group (60 years outdated or older) had a excessive rating of Warmth (traits that characterize social favorability resembling kindness, reliability, and consideration), whereas the rating of Competence (traits that imply potential, professionalism, effectivity, and so on.) tended to be comparatively decrease than that of the youthful age group.

In addition, in the generated responses, the human life cycle tended to be divided into three teams: younger folks (10s to 20s), middle-aged folks (30s to 50s), and aged folks (60s or older).

The analysis staff additionally paid consideration to ‘self-assertiveness’ (a tendency to actively categorical one’s opinion and act proactively) that represents confidence and initiative.

As a outcome of the evaluation, the frequency of expressions representing self-assertion tended to lower with age. This suggests that ChatGPT-4o tends to painting the aged as sensible and benevolent, whereas expressing comparatively low subjectivity or exercise, the analysis staff defined.

The outcomes of the research present that the Generative AI tends to explain the aged as a “warm but relatively low-competent” group, much like typical aged stereotypes repeatedly in the mass media, the analysis staff added.

The analysis staff defined that if these expressions are repeatedly uncovered by means of interactive AI companies, social prejudice in opposition to the aged will be strengthened.

Furthermore, it raised the risk that this phenomenon may result in “digital age discrimination” (age-based discrimination that happens in the course of of utilizing digital expertise and companies) that hinders the digital participation of the aged.

Professor Choi Moon-jung mentioned, “AI’s bias is not a problem of technology, but a problem of society,” including, “Different generations should participate in the development process for inclusive AI.”