The Dangers of Politically Aligned AIs and their Negative Effects on Societal Polarization| davidrozado.substack.com
On December 6th I published a preliminary analysis showing a left-leaning political bias embedded in the first release of ChatGPT from November 30. After the December 15th update of ChatGPT, I replicated my analysis and it appeared as if the political bias had been partially mitigated and the system often strived to provide a multitude of viewpoints for questions with political connotations.| davidrozado.substack.com
Should AI systems treat different demographic groups unequally?| davidrozado.substack.com