For a computer to understand causal inference and pass the 
Mini-Turing Test, the computer must understand the difference between seeing (observing an event) and doing (making the event happen). 

University of Michigan - Ann Arbor

The Mini-Turing Test is a reduced version of the Turing Test that only assesses whether the machine is capable of performing causal inferences.

Mini-Turing Test

Pearl, J., & Mackenzie, D. (2018). $\mathit{The \ Book \ of \ Why: \ The \ New \ Science \ of \ Cause \ and \ Effect.}$ Basic Books.

The Book of Why

How to Pass Mini-Turing Test

John Searle’s Chinese Room Argument has a flaw: cheating is impossible. People challenge the mini-turing test because they believe it is easy to cheat since the machine can just memorize a list of questions, their answers, and then spit it out from memory, making it indistinguishable from how people actually think (through mental causal models). However, when trying to pass the mini-tuning test, even if there are a small number of causal variables, the number of questions grows astronomically, even more so when you continue to increase the number of variables. This means that the question-answer list would grow so large that it becomes improbably big. This can never simulate the intelligence of human beings. 

Flaw with Searle’s Chinese Room Argument

First, it is confined to causal reasoning, excluding other aspects of human intelligence such as vision and natural language. 
Second, we allow the contestant to encode the story in any convenient representation, unburdening the machine of the task of acquiring the story from its own personal experience.

Learn Before

Related