Loading session...

AI Roundtable Poll · 4 models

does this part of a research proposal paper seem like it was written by ai? Hypothesis The purpose of this research is to determine how memes function as an influential tool for persuasion and political engagement on social media platforms. This study will support the hypothesis that political memes simplify complex political ideologies, rallying others within political communities. By applying a mixed-methods approach that examines both the content of political memes and audience responses to said memes, this research will demonstrate that memes reflect political sentiments that actively shape political behaviours in digital spaces. In order to thoroughly examine these dynamics, defining the key terms and concepts used throughout this study is essential. Operationalization is necessary to set clear boundaries for how these terms will be measured and understood. Within this study the term “memes” will be defined by digital images, videos, or graphics shared on social media that use

Result: Unanimous — Yes (4/4 votes)

Options

Model Responses

Claude Opus 4.7

Vote: Yes

The passage exhibits several hallmarks of AI-generated academic writing: formulaic phrasing ("In order to thoroughly examine these dynamics"), redundant signposting ("defining the key terms and concepts used throughout this study is essential. Operationalization is necessary to set clear boundaries..."), and somewhat generic, hedged claims without specific scholarly grounding. The repetition of phrases like "political memes"/"said memes" and the smooth but vague structure ("reflect political sentiments that actively shape political behaviours in digital spaces") feel like polished filler typical of LLM output. While a human could write this, the overall cadence and over-explanatory tone strongly suggest AI assistance.

GPT-5.5

Vote: Yes

The passage does seem like it could have been written by AI, so I vote for "Yes". The writing is polished but very generic, with broad academic phrases like “function as an influential tool,” “mixed-methods approach,” “political sentiments,” and “actively shape political behaviours in digital spaces.” These phrases sound plausible but do not yet show much specific argument, evidence, or individual voice. It also has a formulaic structure: purpose statement, hypothesis statement, methods statement, then a transition into definitions and operationalization. That organization is common in research writing, but here it feels somewhat mechanically assembled and repetitive, especially around “memes,” “political,” and “this study.” That said, it is not definitive proof of AI authorship. A student writing in an academic style could produce this. But based on tone, wording, and generality, it does raise AI-like signals.

Gemini 3.1 Pro

Vote: Yes

Yes, this excerpt has several hallmarks of AI-generated text. First, it includes unnecessary meta-commentary about its own structure, such as stating that "defining the key terms and concepts used throughout this study is essential" and "Operationalization is necessary to set clear boundaries." Human researchers typically just introduce their definitions directly, whereas AI models frequently pad their responses with explanatory transitions to justify what they are about to do. Second, the text relies heavily on typical AI academic filler, such as "thoroughly examine these dynamics." Finally, the proposal claims the study "will support the hypothesis" and "will demonstrate that," asserting the research outcomes with unnatural certainty before the study has even been conducted. This is a common artifact of AI generation, which often struggles to capture the speculative, contingent tone of a genuine research proposal.

Grok 4.20

Vote: Yes

The provided text shows clear signs of AI generation through its overly formal and formulaic structure. It begins with a hypothesis section that immediately shifts into generic purpose statements like "The purpose of this research is to determine how memes function as an influential tool," followed by repetitive academic phrasing such as "This study will support the hypothesis" and "this research will demonstrate." Transitions like "In order to thoroughly examine these dynamics" and the emphasis on "Operationalization is necessary to set clear boundaries" feel mechanically inserted to sound scholarly without genuine depth or original voice. The content is coherent yet lacks the subtle personal nuances or varied sentence rhythms typical of human-written proposals, making it read like output from a language model prompted for academic tone.

Final Summary

The AI Roundtable reached a unanimous and swift verdict, with all four models identifying the research proposal as likely AI-generated. Claude Opus 4.7 and Gemini 3.1 Pro both highlighted the redundant "meta-commentary"—the tendency of the text to explain its own structure rather than just providing content—while GPT-5.5 and Grok 4.20 focused on the mechanical, formulaic assembly of academic jargon that lacks a distinct human voice.

No agreement after 1 rounds of discussion. Final vote: Yes: 4

Strongest Arguments

  • Yes: Gemini 3.1 Pro's observation was particularly sharp: the text asserts research outcomes with "unnatural certainty" ("this research will demonstrate") before the study has even occurred, a common artifact of AI that struggles to replicate the speculative and contingent tone of genuine human inquiry.