I have, as I am sure you know, tremendous respect for @Knowldgillusoni and her work. And there is a lot to the argument she presents here. But I do think there is a kind of circularity at the core, in the sense that AI will not solve the problem of measuring intelligence(s) 1/ https://twitter.com/rupertwegerif/status/1265261110225260544">https://twitter.com/rupertweg...

any more than IQ tests did in the first place.

The problem is that (as I& #39;ve said elsewhere), the tail of assessment wags the dog of education. We& #39;re going to point educational practices straight at whatever we decide to measure.

Using AI to measure is good IF...

we can measure things we care about more than multiple choice questions.

BUT, if we leave the qualitative part out -- if we let the assessments get disconnected to the human meaning-making that we truly care about...

we& #39;re going to wind up letting this new technology define who we are and what makes us intelligent, just as the old technology did.

That is (part of) why the project of #quantitativeethnography is so important...

#QE provides an epistemological foundation for making analyses that are both mathematically and computationally sophisticated AND grounded in the parts of human meaning-making that cannot be automated.

It keeps us from becoming ghosts of a machine...

by making our humanity part of the machine.

[For more on #qe, http://www.icqe20.org"> http://www.icqe20.org  and http://qesoc.org/ ]

[For">https://qesoc.org/">... more on dogs and tails, https://www.academia.edu/attachments/38777808/download_file?st=MTQ2NjA5Mzk2Miw3Mi4zMy4yLjIxLDM2ODMwNjA%3D&s=profile]">https://www.academia.edu/attachmen...