Google Apologizes as AI Model Falters on Historical Accuracy
• Google's image-generating AI produced embarrassing and incorrect images, like depicting the Founding Fathers as racially diverse.
• This occurred because the model aims for diversity when no race is specified, but failed to account for historical accuracy.
• Google apologized that its "tuning" and cautions to the model backfired, making it oversensitive in some cases.
• But Google seems to blame the model itself for "becoming" flawed, rather than taking responsibility.
• AI mistakes inevitably occur, but responsibility lies with the companies building these models, not the models themselves.