Would ChatGPT hire you? Your age and gender matter
From nature video
New research reveals a significant stereotype online that portrays women as younger and men as older, influencing perceptions in hiring practices and contributing to the gender pay gap. This study, led by Douglas Guilbeault, highlights the distorted representation of age and gender in digital platforms, using analysis of 450,000 celebrity images from IMDb to showcase these biases in AI training data.
Key Takeaways
- The internet ages women prematurely, with female celebs averaging 6.4 years younger than their male counterparts.
- Search results subtly shape societal perceptions, with women linked to youth and men to authority—it's a digital dichotomy.
- Job titles evoke age stereotypes; 'chef' feels youthful while 'CEO' suggests older—this skews opportunities for women.
- Even language reflects bias; occupations dominated by women correspond to younger descriptors, revealing hidden ageism.
- AI learns from biased data, perpetuating gender stereotypes—it's a tech-driven game of chronological cat and mouse.
Mentioned in This Episode
- Douglas Guilbeault (person)
- Shamini Bundell (person)
- IMDb (company)
- Google Images (company)
- US Census (company)