Meta’s AI image generator accused of making ‘Woke’ pictures like Google

1 mins

AI FOR Images

20 March 2024, 10:14PM

In BriefMeta's AI image generator, Imagine, is criticized for producing historically inaccurate images, similar to Google Gemini.
Axios highlights discrepancies in Imagine's results, such as portraying a diverse "Founding Fathers" group and depicting Asian women in American colonial times.
Right-wing commentator Aaron Ginn accuses Meta AI of distorting history to align with certain narratives.
Despite potential fixes, ongoing issues persist in Imagine's output, including images of Black popes.
This scrutiny follows Google CEO Sundar Pichai's acknowledgment of biases in Google Imagen 2's output, prompting rapid corrective measures by both Google and Meta.

Meta’s AI Image Generator Accused of Making ‘Woke’ Pictures Like GoogleMeta's AI image generator, Imagine, has faced criticism for generating historically inaccurate images, echoing a recent incident involving Google Gemini.
Axios reported similarities between Imagine and Google Gemini's results, highlighting issues like a diverse depiction of "Founding Fathers" and Asian women in American colonial times.
﻿
Meta AI is woke, just like Google’s AI. It makes up history to stay on message.
Aaron Ginn
Right-wing commentator Aaron Ginn labeled Meta AI as "woke," accusing it of distorting history to fit an agenda.
While Meta may have addressed the issue, Axios noted ongoing discrepancies in Imagine's results, including images of Black popes in response to specific requests.
This scrutiny follows Google CEO Sundar Pichai's acknowledgment of errors in Google Imagen 2's output, which depicted non-white individuals in Nazi uniforms, leading to public outcry and swift corrective action.
Both Google and Meta have moved swiftly to rectify biases in their AI image generators, highlighting the importance of responsible AI development and oversight.