hellowthis study you refer to published in Nature is the most...

hellow

this study you refer to published in Nature is the most relevant and interesting piece seen on these pages for the future of smart AI and Appen. In a nutshell, but crudely put, AI is only as good as its programmers, wherein programmers includes all the relevant data input. But data and most importantly, its interpretation, is not static and full of nuance beyond even human understanding most of the time, neveralone AI. So as much as AI can do the hard yards in crunching data in a quasi-human way quickly and reasonably acceptable in a bulk of circumstances, it cannot always get the context right. Therefore, any AI has to be smart enough to almost make human analysis obsolete, which of course is only a pipe dream, leaving any AI technology reliant on the human factor - requiring a human soul so to speak - if they are going to be in the forefront of the industry.

Appen, in my first readings seems to be in a good place.

Thanks for the article.