Can AI just replace me already? – A comparative AI-writing ID experiment
The article discusses an experiment conducted by Richard Hanania to test AI's ability to imitate his writing style. Hanania compared AI-generated articles with his own to see if readers could distinguish between them. The results suggest that AI can effectively replicate human writing styles, raising questions about biases in perceptions of AI-generated content.
- ▪Richard Hanania conducted an experiment to see if AI could imitate his writing style.
- ▪He found that AI-generated articles were often indistinguishable from his own by readers.
- ▪The experiment raises questions about biases in how people perceive AI-generated content.
Opening excerpt (first ~120 words) tap to expand
EssaysCan AI Replace Me Already?Results of an AI Turing test survey: Women are ready for our new reality, old people aren'tRichard HananiaMay 17, 2026703711ShareI recently read an article by Nabeel Qureshi on what makes AI art bad. The whole time, I was responding “Yes, but…” He would say great art does X, but AI does Y, which is why AI can’t produce great art. But how do we know we are not biased in our judgments? Maybe you just see stuff made by AI and assume it’s bad. We need blind testing, or such observations are not worth much. When Scott Alexander did this in 2024, he found that people could differentiate AI and human art 60.6% of the time, which isn’t very impressive when you consider that you would get 50% by chance.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Hacker News (AI / LLM).
Discussion
0 commentsMore from Hacker News (AI / LLM)
