Teach AI to recognize flawed arguments before trusting them.
AI systems are increasingly used to summarize debates, assess claims, and surface information for decision-makers. But models trained on human text inherit human reasoning patterns, including the flawed ones.
Logical fallacies undermine arguments in subtle ways that AI doesn't naturally detect. Teaching models to spot these errors requires human judgment.
The Logical Fallacy Detection Quest is now live on Perle Labs! Visit app.perle.xyz to start earning points.
What This Task Is
The Logical Fallacy Detection Quest is a set of 10 multiple choice tasks. Each prompt presents a short argument from debates, opinion pieces, or everyday discussions, and each argument commits a logical fallacy.
Your job is to identify which type of fallacy is present. These classifications help train AI systems to recognize flawed reasoning, improve fact-checking capabilities, and better understand how humans construct (and misconstruct) arguments.
What You'll Do
Step 1: Read the argument carefully.
Step 2: Identify the flaw in reasoning that makes the argument invalid or weak.
Step 3: Select the option that best identifies the type of fallacy present.
Step 4: Submit your answer and move to the next task.
Each passage contains one primary logical fallacy. Choose the single most accurate answer.
Fallacy Types
- Ad Hominem: Attacking the person rather than the argument
- Straw Man: Misrepresenting an argument to make it easier to attack
- False Dilemma: Presenting only two options when more exist
- Slippery Slope: Claiming one event inevitably leads to extreme consequences
- Appeal to Authority: Using an authority's opinion when they lack relevant expertise
- Circular Reasoning: Using the conclusion as a premise
- Red Herring: Introducing an irrelevant topic to divert attention
- Appeal to Emotion: Using emotional manipulation rather than logic
- Hasty Generalization – Drawing broad conclusions from limited examples
- False Cause (Post Hoc): Assuming sequence implies causation
Why It Matters
By contributing to this task, you are:
- Teaching AI to detect reasoning errors in arguments and claims
- Building enterprise-grade datasets for fact-checking and content analysis systems
- Improving how AI evaluates the validity of information it processes
- Contributing to the sovereign data layer that enterprises and institutions require
This is critical thinking work that AI cannot learn alone, the human layer that makes information systems more reliable.
Task Details
- Points per question: 100
- Number of tasks: 10
- Total points available: 1,000
- Project eligibility: 1,000 points
Visit the app today to start earning points!
Ready to Get Started?
The Logical Fallacy Detection Quest is now live on your Perle Labs dashboard: http://app.perle.xyz.
Each classification you submit helps build auditable AI infrastructure for systems where trusting flawed reasoning isn't an option, not black-box pipelines, but human-verified data with a clear chain of custody.
Stay Connected
Whether you’re returning from beta or joining for the first time, Season 1 is open and ready. Follow along with us on X, Discord, and Telegram to stay connected so you don’t miss what’s coming.
Human-verified. Auditable. Sovereign. Start contributing.
