5 ms·

Scientists Invented a Simple New Test for Creativity, and You Can Try It Out

12 points by elorant 5y ago

lonnydonovan 5y ago
Here's the thing. Creativity is measured by a subjective human mind. If you can simulate the mind's current state using AI (with relevant inputs), you can probably measure creativity. For some reason, we humans glorify certain things like creativity, attributing those to some insurmountable human ability. As a scientist, I don't find anything extraordinary in this. Just a simple ability to create. After all, who is is the judge? Ourselves!
- audit 5y ago
  it seems that 'creativity' measure should be based on the judgment of others who presumably, spend a lot of time trying to solve similar problems. If creativity, is just a measure of 'surprise' -- then it could be easily subjected to unfair measurement. Because 'surprise' on its own can happen easily.
Anunayj 5y ago
For anyone interested, what the study does is, asks you to name 10 different nouns, and then searches for the semantic distance (how often they are not used together) between them to give you a score. So I just fed it a list of computer generated random nouns with 2 syllables each and got a 95th percentile.
- mousepilot 5y ago
  Thats pretty creative!
lennardv 5y ago
Score of 90. I doubt this really invokes creativity (in my case). I just start making a vowel sound in my head and then finish it with a word, so its more sound oriented than based on association.
- gattilorenz 5y ago
  But that's a creative solution to finding random words :)
xyzzy123 5y ago
It seems like there's an interesting bias in the test methodology. If you are a person who prefers simple, common or short words, the test is much harder because short words tend to have many distinct meanings. When you cross correlate a lot of short words there will inevitably be semantic overlap. Longer, more specialised words have greater semantic distance. Therefore not clear whether this measures creativity or propensity to use longer / less common words. This will tend to correlate with "years of formal education". It would be interesting to see if this is measuring anything different than a vocabulary test + the ability to understand the instructions. One option to balance this out would be to "weight" words by their frequency to reflect that finding disjoint common words is harder.
audit 5y ago
Here is my result (first attempt): --- "... Your score is 81.1, higher than 67.85% of the people who have completed this task... The average score is 78, and most people score between 74 and 82. The lowest score was 24 and the highest was 96 in our published sample. Although the score theoretically ranges from 0 to 200, scores above 100 are extremely unlikely on first attempts. The highest total score we have seen is 107. See how other people are performing on Twitter. ..." --- Second attempt was 70 something, --- Third attempt was: ".. Your score is 84.13, higher than 82.07% of the people who have completed this task ..." It looks like I cannot break 85. I am not a native English speaker, but very fluent...

12 points by elorant 5y ago

Here's the thing. Creativity is measured by a subjective human mind. If you can simulate the mind's current state using AI (with relevant inputs), you can probably measure creativity. For some reason, we humans glorify certain things like creativity, attributing those to some insurmountable human ability. As a scientist, I don't find anything extraordinary in this. Just a simple ability to create. After all, who is is the judge? Ourselves!

audit 5y ago
it seems that 'creativity' measure should be based on the judgment of others who presumably, spend a lot of time trying to solve similar problems. If creativity, is just a measure of 'surprise' -- then it could be easily subjected to unfair measurement. Because 'surprise' on its own can happen easily.

Anunayj 5y ago

For anyone interested, what the study does is, asks you to name 10 different nouns, and then searches for the semantic distance (how often they are not used together) between them to give you a score. So I just fed it a list of computer generated random nouns with 2 syllables each and got a 95th percentile.

mousepilot 5y ago
Thats pretty creative!

lennardv 5y ago

Score of 90. I doubt this really invokes creativity (in my case). I just start making a vowel sound in my head and then finish it with a word, so its more sound oriented than based on association.

gattilorenz 5y ago
But that's a creative solution to finding random words :)

xyzzy123 5y ago

It seems like there's an interesting bias in the test methodology. If you are a person who prefers simple, common or short words, the test is much harder because short words tend to have many distinct meanings. When you cross correlate a lot of short words there will inevitably be semantic overlap. Longer, more specialised words have greater semantic distance. Therefore not clear whether this measures creativity or propensity to use longer / less common words. This will tend to correlate with "years of formal education". It would be interesting to see if this is measuring anything different than a vocabulary test + the ability to understand the instructions. One option to balance this out would be to "weight" words by their frequency to reflect that finding disjoint common words is harder.

audit 5y ago

Here is my result (first attempt): --- "... Your score is 81.1, higher than 67.85% of the people who have completed this task... The average score is 78, and most people score between 74 and 82. The lowest score was 24 and the highest was 96 in our published sample. Although the score theoretically ranges from 0 to 200, scores above 100 are extremely unlikely on first attempts. The highest total score we have seen is 107. See how other people are performing on Twitter. ..." --- Second attempt was 70 something, --- Third attempt was: ".. Your score is 84.13, higher than 82.07% of the people who have completed this task ..." It looks like I cannot break 85. I am not a native English speaker, but very fluent...