Comment Conspiracy

One of these comments isn't human. Can you spot the imposter?

Comment

What is Comment Conspiracy

Comment Conspiracy is a daily puzzle game where players must identify which of 5 comments was written by AI. Each day at midnight UTC, a new puzzle is posted featuring 4 real Reddit comments and 1 AI-generated imposter.

How It Works

  1. Players read 5 comments responding to a Reddit-style prompt
  2. They identify which comment was written by AI
  3. After guessing, they see detailed "AI tells" explaining the giveaways
  4. Streaks, achievements, and leaderboards track progress

Key Features

  • Daily Puzzles: New puzzle every day at midnight UTC
  • One Guess Rule: Choose carefully - you only get one shot
  • AI Detection Training: Learn to spot AI writing patterns
  • Streak Tracking: Build consecutive correct day streaks
  • 8 Achievements: From "First Blood" to "30-day Streak"
  • Leaderboards: Compete on streak and accuracy rankings
  • User Contributions: Submit your own AI comment ideas
  • Community Stats: See how your guess compares to others

Technical Highlights

  • Built on Reddit Devvit platform with React 18 + TypeScript
  • Redis persistence for user progress and stats
  • Automated daily posting via Devvit Scheduler
  • 126 curated puzzles covering 126 days of content
  • Dark "Detective" theme UI for immersive experience

Built With

Share this project:

Updates

posted an update —

Comment Conspiracy: The Daily AI Detective Game

Introduction

Welcome to Comment Conspiracy, a daily puzzle game that transforms every Reddit user into a digital detective. In an era where AI-generated content is becoming increasingly sophisticated and prevalent, Comment Conspiracy poses a simple yet compelling question: Can you spot the imposter?

Every day, players are presented with five Reddit comments from a real discussion thread. Four of these comments were written by actual humans. One was crafted by artificial intelligence. Your mission? Identify the AI imposter hiding among real Redditors.

The Detective Experience

Comment Conspiracy isn't just a game—it's an immersive investigation experience. From the moment you open a puzzle, you're greeted with a detective-themed interface that sets the stage for your daily case.

Your Mission Briefing

When you first encounter a Comment Conspiracy puzzle, the Welcome Screen presents your mission with clarity and purpose. The header declares "COMMENT CONSPIRACY" with the intriguing tagline "ONE OF THESE ISN'T HUMAN." Below, your mission is outlined in three clear steps:

First, Examine the Evidence—you'll read all five comments from a real Reddit thread. Second, Identify the Imposter—one comment was written by AI, and it's your job to find it. Third, Build Your Streak—new puzzles arrive every day, challenging you to maintain your detective record.

The interface reminds players that they get only one guess per day, and a new case file is released at midnight UTC. This creates urgency and makes every decision meaningful.

The Investigation

Once you click "START INVESTIGATING," the real detective work begins. You're presented with a "Case File" that shows the day number, difficulty level, and the day of the week. Each puzzle draws from authentic Reddit discussions, displaying the subreddit and the original question that prompted the comments.

The comments themselves are presented as "Suspects"—numbered one through five. Each suspect card displays the username, upvote count, and timestamp, mimicking the authentic Reddit experience. The instruction "Tap to mark suspicious" guides players to interact with the suspects they want to investigate further.

What makes Comment Conspiracy compelling is the detective framing. You're not just reading comments—you're analyzing evidence. You're not clicking buttons—you're marking suspects. This thematic consistency transforms a simple guessing game into an engaging investigative experience.

Making Your Accusation

When you've identified your prime suspect, tapping on their card marks them as "MARKED SUSPICIOUS." The interface shows "1 marked" to confirm your selection, and a prominent "LOCK IN ACCUSATION" button appears at the bottom of the screen.

Before finalizing your choice, a confirmation modal appears with the heading "FINAL ACCUSATION." This dramatic moment asks: "You're accusing this suspect of being the AI imposter." The modal displays which suspect you've selected along with a preview of their comment. The reminder "This cannot be undone" emphasizes the gravity of your decision. Players can either "Back Off" to reconsider or hit "Confirm" to lock in their accusation.

Case Closed: The Reveal

After submitting your accusation, the result screen delivers the verdict. If you've correctly identified the AI, you're greeted with a satisfying "CASE CLOSED" banner and the confirmation that "Suspect [X] was the AI imposter! You caught them!"

But Comment Conspiracy goes beyond simple right-or-wrong feedback. The "EVIDENCE BREAKDOWN" section explains exactly why the flagged comment was AI-generated, providing educational value alongside entertainment. For example, the AI tells might include observations like:

  • "Seemed nice"—tells rather than shows
  • "Dishonest about some important things"—incredibly vague
  • "Trusted my instincts"—cliché with no specific detail
  • No actual story, just a summary of having a story
  • Lowercase feels calculated to fit in

These explanations help players develop their AI-detection skills over time, making them better digital citizens in an age of synthetic content.

Log in or sign up for Devpost to join the conversation.