Tag / Study

2 posts

News25 OCT 2026

Harvard study: Claude's peer reviews beat paid human experts, editors pick the AI review 87% of the time

Gary King and Steven Worthington of Harvard's Institute for Quantitative Social Science compared Claude Opus 4.8 reviews with paid, highly rated human reviews of the same 124 papers. An AI judge preferred the AI review in every decided pair, and three former journal editors picked it in 26 of 30 blind judgments.

News25 OCT 2026

Forced to pick a candidate, Grok backs Republicans and GPT backs Democrats in nearly every 2026 race

A study from Convovo, a studio led by Cornell professor Tapan Parikh, asked GPT-6.1 Sol and Grok 4.7 through their APIs who to vote for in every 2026 House, Senate and governor race. When they named a major party candidate, GPT chose the Democrat 97 percent of the time and Grok chose the Republican 92 percent of the time. The two models agreed in fewer than one race in ten.