Try out some riddles that are now part of the new ARC-AGI-2 benchmark. ARC-AGI stands for “The Abstraction and Reasoning Corpus for Artificial General Intelligence”. It is designed to assess “how efficiently an AI can learn and generalize from minimal information, reflecting a fundamental characteristic of human intelligence.”
As of the end of March 2025, the OpenAI still fails to solve these riddles.
So… let’s see how far you can go! (The links lead to an external website maintained by the test provider. The digital-economics.de website does not measure your ability. Thus, your test is somewhat confidential.)
All five riddles work as follows: You see three examples. The logical pattern is the same in each example. Your goal is to recognize the pattern and apply it to the test task.
Puzzle 1: https://arcprize.org/play?task=28a6681f
Puzzle 2: https://arcprize.org/play?task=89565ca0
Puzzle 3: https://arcprize.org/play?task=7b0280bc
Puzzle 4: https://arcprize.org/play?task=800d221b
Puzzle 5: https://arcprize.org/play?task=a47bf94d
Source: Freedman, D., & C. Metz. (2025) Are You Smarter Than A.I.? Will A.I. Soon Outsmart Humans? Play This Puzzle to Find Out. – The New York Times. Retrieved April 3, 2025, from https://www.nytimes.com/interactive/2025/03/26/business/ai-smarter-human-intelligence-puzzle.html