[peekyboy]

what will education reform look like in the next 100 years now that AI and ML have ripped apart the standardized tests ?

[Dismas]

There will probably more emphasis on comprehension and understanding of material. Essays always seemed more focused on length/regurgitation. Certain subjects will probably be more vulnerable to AI. For subjects with less complex content, professors need to tweak questions to be more specific.

They are already using AI to detect AI-written papers, so there will be countermeasures similar to Turnitin. It won't stop plagiarism or AI-written papers, but I'm sure professors will be more likely to base grades on other mediums. Most degrees require in-person exams, presentations, and stuff an AI can't do.

[mothered]

Moved from Random Discussion to Academic Discussion.

[HackingRealm]

Moved from Random Discussion to Academic Discussion.

I never saw academic discussions section on forum till new. Lol. I should explore to see other sections i never visited.

[mothered]

I never saw academic discussions section on forum till new. Lol.

All good.

It takes around 8 seconds to move a thread.

[ZeusRTW]

What will things such as neuralink have on the coming years of education?

[panarat203]

School definitely need to be more rigorous, maybe paper only testing, while childhood education should be teaching kids to be non-cheating responsible students to prepare for university.

[BigSuc]

There will probably more emphasis on comprehension and understanding of material. Essays always seemed more focused on length/regurgitation. Certain subjects will probably be more vulnerable to AI. For subjects with less complex content, professors need to tweak questions to be more specific.

They are already using AI to detect AI-written papers, so there will be countermeasures similar to Turnitin. It won't stop plagiarism or AI-written papers, but I'm sure professors will be more likely to base grades on other mediums. Most degrees require in-person exams, presentations, and stuff an AI can't do.

I've been thinking recently about how somebody's identity can be inductively found through analyzing their written text. If a simple cipher text is long enough, for instance, frequency analysis can be used to easily break it because the the same per-letter frequency rate shows up in roughly all English text. Similarly, you could imagine each word of a sentence having a relative frequency compared to the whole text. People have an individual preference for certain sentence structures, or turns of phrase, or just vocab choice. Given two words that occupy the same space, like "everyone" vs "everybody", someone might choose to use one consistently over the other.
Once you know someone's verbal fingerprint, you might be able to compare it to ChatGPT and see how similar they are, and probably use some statistics 101 to tell how significant the similarities and differences are. Scary shit