- “Is the definition of intelligence we use for humans
applicable to machines?”
- “Have machines (AlphaGo, IBM Watson, Viv) already surpass human intelligence?”
- “How can we build a decision boundary to measure Human and Machine intelligence?”
To answer some of the question poised above, I have thought of a simple test similar to Turing test. Alan Turing believed machines are intelligent enough if they pass turing test. But in recent years, chatbots like Goostman have fooled some human judges in believing them as humans. So, does it mean that these chatbots are already in level with human intelligence ? I don’t think so. I purpose a new test to make a decision boundary between machine and human intelligence.
Nobel Prize Test
Nobel prize was established according to the will of Alfred Nobel to recognize and accolade the one for outstanding contributions for humanity in chemistry, literature, peace, physics, physiology or medicine.
A single act of a person can change others life forever. One simple discovery of a scientist can save million lives. Humans have an extraordinary power to influence others. We are directly or indirectly influencing the life of every living being on this planet. This influential power is what separates us human for rest of the animal kingdom. Based on this influential power, we can compare the intelligence of human and machine. I purpose a test where this influential power will be the key metric to compare human and machine intelligence.
- If a machine can discover a nobel phenomena (Netwon’s Gravitation Theory, Einstein theory of Relativity, Faraday’s Law) and publish it to top tier peer reviewed journals, then machine can be considered as equal to human intelligence.
- If a machine can grab awards like Nobel Prize, Fields Medal or Turing Award for outstanding contributions for humanity, then machine can be considered above human intelligence.
How does the test mentioned above justify the comparison of intelligence between Human and Machines ?
I will explain in great depth in my next post on how the above test would be a good metric to distinguish between human and machine intelligence.