Abstract or AI? A Researchers’ Guessing Game

As with the rest of the world right now, artificial intelligence (AI) is rapidly transforming research, and sometimes the lines between genuine studies and AI-generated content are becoming increasingly blurred.
How well can you distinguish between cutting-edge science and imaginative machine output?
This quick quiz invites academics, students, and research enthusiasts to put their skills to the test in a fun and thought-provoking way.
No Googling allowed, rely solely on your research instincts.
And no peeking below: the answers are at the end of the page.
Can you tell which of these research abstracts are real and which were generated by AI?
Abstract 1:
Neural Correlates of Memory Encoding in Urban Crows
Crows are known for their problem-solving skills, but little is known about their memory encoding at a neural level. This study investigates the hippocampal activity in urban Corvus brachyrhynchos during a spatial navigation task using portable EEG tracking. Findings suggest that environmental complexity influences memory consolidation patterns in these birds.
Is this real or AI?
Abstract 2:
The Effect of Caffeine on Quantum Fluctuation Observation Among Graduate Physicists
This study evaluates whether increased caffeine intake enhances physicists’ ability to detect quantum fluctuations during late-night lab sessions. Results show a statistically significant correlation between espresso consumption and perceived particle-wave duality, though causation remains unconfirmed.
Is this real or AI?
Abstract 3:
Socioeconomic Impacts of Water Scarcity on Rural Farming Communities in the Eastern Cape
This paper analyses the long-term economic and social effects of water scarcity on small-scale farmers in South Africa’s Eastern Cape province. Using a mixed-methods approach, the study found that reduced rainfall and inadequate irrigation infrastructure significantly hinder agricultural output and household resilience.
Is this real or AI?
Abstract 3:
Machine Learning Applications in Predicting Romantic Compatibility Based on Smell Preferences
Combining olfactory profiling with supervised learning algorithms, this exploratory study seeks to determine whether romantic compatibility can be predicted based on scent preferences. Participants underwent blind scent trials followed by matchmaking surveys, with moderate predictive success reported.
Is this real or AI?
Answers:
1
REAL – Inspired by actual research on avian cognition and memory.
2
AI-GENERATED – As fun as it sounds, this caffeine-fuelled quantum study is purely fictional.
3
REAL – Drawn from a genuine study examining the impacts of water scarcity on Eastern Cape farmers.
4
AI-GENERATED – Not a real study (yet), though we’d be curious to see it come to life.
—
So, what have we all learnt?
In this exciting but often uncertain age of AI, distinguishing authentic research from generated content has become a critical skill for the academic community.
This exercise is not only entertaining but also a subtle reminder of the importance of critical thinking, careful analysis, and verifying sources.
Whether you aced the quiz or learned something new, the game underscores a key truth:
Curiosity and discernment remain the most reliable tools in a researcher’s toolkit.