This is a GREAT post - though my knowledge of fluency is different to yours?
I totally understand exactly what your argument about improving reading is - because checking AI Results for errors is EXACTLY what we all should do! And agreed - it does SLOW reading - and that’s because it changes the reading goal - which isn’t JUST to read the content - the goal has become - I need to check that AI text to see if it’s an error?
THANKS so much for raising and sharing an important point about reading AI texts! 👍
I think that’s definitely part of it — checking for errors is one piece of the shift.
Where I’d extend things is that even when there aren’t clear factual errors, that doesn’t mean the argument itself holds up. A lot of what I’m seeing isn’t outright wrong so much as more superficial — the structure is there, it reads well, but the depth and nuance you’d expect from sustained engagement just aren’t. This will become more patent in Part 3 of the series.
So the slowdown isn’t just about verifying accuracy. It’s also about having to dissect the reasoning — what’s being assumed, how the pieces fit together, whether anything substantive is actually happening underneath the prose.
That part becomes more obvious the more the substance gets offloaded.
Thanks Matt, and I agree with your points here, again! What you describe (beyond accuracy of both decoding and facts) is exactly what has been discussed as "deep reading comprehension in reading research...
And that term "deep reading comprehension" has been defined in many ways - in my doctoral research (a while ago) in Question-Answer Relationships, my comment about accuracy was more focussed on Right There (Literal) QAR - in your reply, this is the surface / superficial level. The other types are "Think & Search" & "On My Own" - this is the language used in my research for elementary students...
For me, when teaching younger students, this language helps them understand where answers might come from - and it's NOT always the text.
It's these "higher order" cognitive processes you describe as "substantive", how concepts fit together and underlying inferences - these are what students find most difficult to learn, I found?
Hope this makes sense, in your thinking - as I see a clear connection between these ideas and your current post. I look forward to reading more of what you think in your next posts! MANY THANKS!
This is a GREAT post - though my knowledge of fluency is different to yours?
I totally understand exactly what your argument about improving reading is - because checking AI Results for errors is EXACTLY what we all should do! And agreed - it does SLOW reading - and that’s because it changes the reading goal - which isn’t JUST to read the content - the goal has become - I need to check that AI text to see if it’s an error?
THANKS so much for raising and sharing an important point about reading AI texts! 👍
I think that’s definitely part of it — checking for errors is one piece of the shift.
Where I’d extend things is that even when there aren’t clear factual errors, that doesn’t mean the argument itself holds up. A lot of what I’m seeing isn’t outright wrong so much as more superficial — the structure is there, it reads well, but the depth and nuance you’d expect from sustained engagement just aren’t. This will become more patent in Part 3 of the series.
So the slowdown isn’t just about verifying accuracy. It’s also about having to dissect the reasoning — what’s being assumed, how the pieces fit together, whether anything substantive is actually happening underneath the prose.
That part becomes more obvious the more the substance gets offloaded.
Thanks Matt, and I agree with your points here, again! What you describe (beyond accuracy of both decoding and facts) is exactly what has been discussed as "deep reading comprehension in reading research...
And that term "deep reading comprehension" has been defined in many ways - in my doctoral research (a while ago) in Question-Answer Relationships, my comment about accuracy was more focussed on Right There (Literal) QAR - in your reply, this is the surface / superficial level. The other types are "Think & Search" & "On My Own" - this is the language used in my research for elementary students...
For me, when teaching younger students, this language helps them understand where answers might come from - and it's NOT always the text.
It's these "higher order" cognitive processes you describe as "substantive", how concepts fit together and underlying inferences - these are what students find most difficult to learn, I found?
Hope this makes sense, in your thinking - as I see a clear connection between these ideas and your current post. I look forward to reading more of what you think in your next posts! MANY THANKS!
can confirm, but then I never wrote smoothly, and this can be styled into our algorithms for flowers too