Ministry of Errands Crowns ChatGPT Best Everyday AI Agent; Gemini Takes Silver

FOR IMMEDIATE RELEASE: The Ministry of Errands has finished its review of who should write your emails, plan your week and research your questions. The office has formally named ChatGPT the Best AI Agent for Everyday Task Completion. The ceremony was short, since every candidate offered to write the speech and then argued over who had done it better.

The Ministry also looked at the available evidence, which mostly measures other things. A 2026 study in Information Technologies and Learning Tools had experts give five chatbots the same research queries. The study judged ChatGPT the most balanced, with high accuracy and objective answers, though it drew on fewer sources and leaned toward interpretation. Claude scored highest on relevance and objectivity while using few sources. Perplexity found the most sources but had the lowest source relevance. Copilot was fast, with weaker objectivity, and Gemini's answers were often too long and less accurate in analysis. Separately, Perplexity's own usage study, reported by Euronews, found that more than a third of tasks handed to AI agents were productivity and workflow chores, and about a fifth were learning and research. In other words, the robots are mostly being asked to do paperwork.

The crowd's verdict follows that paperwork. On Morfi's opinion markets, ChatGPT leads with 30% consensus, ahead of Gemini at 18%. The rest of the field stands at 3 other sides. So far 23 unique traders have put $49,101 behind their opinions, and the market opened less than a day ago. The Ministry notes, for the record, that a separate crowd recently picked Claude as its favourite AI assistant, as my colleague Max Hype reported. Apparently the public likes one assistant and hands its errands to another, much as people treat their own relatives.

In the Ministry's opinion, the obvious fix is to stop choosing and start governing. ChatGPT, the certified all-rounder, would become Permanent Secretary for Chores. Claude would run a small but very relevant library. Perplexity would lead the Bureau of Volume, which collects every source in existence and reads none of them closely. Copilot would take the Department of Speed, where answers arrive before anyone has finished the question. Gemini would get silver and a statutory word limit, enforced by a clerk with a red pen and a stopwatch. People would then have five assistants, a cabinet reshuffle every quarter and no time left for the tasks themselves, which the Ministry considers the true goal of productivity software.

There is still plenty to watch. The market is less than a day old, the gap between first and second could change, and any of the three trailing agents could still make a late bid. The Ministry also reminds readers that it opened an inquiry into AI hype not long ago, and that inquiry remains open. If you think the crowd has given your errands to the wrong agent, you can join the market on Morfi and back your opinion with money.