This study was shared on ley a few weeks ago, with a less clickbait title:
“Artificial Intelligence”
The problem, of course, is that there is no agreed-upon definition of what “Intelligence” even is. This seems to me to be a critical failure of people wanting to make an artificial version of it. “We’ve made Artificial Geflugenschurz!” “What’s Geflugenschurz?” “That’s not important. What’s important is that we’ve made the artificial version of it!”
There are many determinants of what’s not intelligence though. For example a computer program that just processes data is generally not considered intelligent, no matter how much data it can process.
I’m not sure how that’s relevant.
There’s many determinants of what isn’t a piece of steel. Trying to make, however, a piece of steel by enumerating each item that isn’t a piece of steel is not going to be productive. Generally you have to know what steel is and how it’s made to make one.
We don’t know what an intelligence is. We thus don’t know how to make one.