Gary Marcus

Afaik, essays have also been graded by machines, feedback given and when asked, students could not tell the difference between a mechanized TA and a real person. This was a watershed moment in machine assessment.

And that's, I think, why ISI wants to remove bias. Though I think they would also want to make their #ML fair.

It's a worthy effort, but not reliable until proven. And given current evidence, not necessarily replicable.

Right now, IMHO, knowing current progress in #NLProc, #AI has what you'd call "reading comprehension disability" in children (if I dare anthropomorphise machines just for this once, against @j2bryson's advice ) and that's not getting fixed soon enough for @GaryMarcus to relent.

That large proportions of people have reading skills lower than high-end NLP a) has been demonstrated by the @oecd b) hasn’t been taken adequately into account by the #agi #singularity #Futureofwork panic crowds.

more generally @j2bryson what does your tweet even mean? is there a typo? what OECD study? what was the metric? what implications? having a hard time making sense of what you said.

The limit case is obvious – some people are fully illiterate so lots of AI gets more from text than they do. I'm illiterate in Chinese, for example. The OECD at their 2017 AI meeting talked about a study that showed that 70% of EU citizens can't do analyses top-end AI can 1/2