Results are in: 80% of readers took "solving a Rubik's cube" to refer cognition & manipulation, rather than manipulation on its own.
Not a perfect scientific sample, but consistent w fact that many were surprised to discover solving in that sense was not in fact learned @openAIhttps://twitter.com/GaryMarcus/status/1185918148916076547 …
-
"Let's poll my followers, who probably follow me because they agree with me, on whether they agree with me."
I noted that it was not scientific, and also posted the excellent independent review by @skynet_today that converge on the same place. If you seriously think you would get radically different results from a random sample of North American population, go out and collect the data.
