yehudalouis

A good friend and I have been discussing the growing feeling that we're ready to "leave" the Internet as we know it. The AI slop on the websites we used to find the most utility in is getting unbearable. While I know that there are still high-quality websites hidden in the corners of the Internet untouched by AI (or relatively so), the fatigue we feel is certainly real.

It got us thinking -- if someone were to design a new Web, a new protocol that requires verification that a human is accessing it and only a human is submitting content/interacting/reading, what would the verification process look like?

Parameters for design: - Only humans can access - Only humans can interact (all CRUD must be done by a human) - Anonymity is the default -- if you wish to identify yourself, feel free - Must be able to use existing technology (a web browser)

What would the initial and ongoing realtime verification look like? How would you design it?

I believe that for most simple writing assignments in K-12, ChatGPT is pretty adept at spitting out a result that would yield a somewhat decent grade.

If you were a teacher, given what you know about ChatGPT, among other tools, and how it will become more and more proficient, how would you structure your classroom to sidestep the concern?

I would probably lean towards in-class debates, where students have a day or two to prepare, but push the topics into more philosophical territories, rather than debates using just simply facts. Thoughts?