Writing is as much about organizing your thoughts and learning how to build an argument as it is about putting words on the page. It is about learning how to think. If students rely on an LLM, they will never get a chance to practice this essential skill, and, in my opinion, will be a lot dumber as a result.
My thoughts are along similar lines: People learn to think, to inquire, to persuade, etc. We don't know precisely how. Maybe it's correlated to education, but even the strength of a causal link is debatable.
I certainly don't oppose reforming contemporary education, but at the same time, letting it be replaced by something else by default because that thing has exponentially more engagement power invokes Chesterton's Fence. I'm not sure we even know what we're giving up.
Education has been in dire need for reform a long time. The writing was on the wall even with the advent of the Internet. We now have supercomputers in our pockets and hyperintelligent assistants at our fingertips.
Chesterton could certainly not predict the exponential growth in access to tooling and knowledge. And we're only at the start of the curve.
In my opinion we should go back to absolute basics. Forget grades. Focus on health, language, critical thinking, tool usage, and creativity. Skills that are intrinsic to humans.
Make sure education is fun with lots of play. The main advantage humans have over computers is empathy and creativity. I'm not sure AI will ever "get it".
Provide each student to the extent possible a path to follow their own curiosity and talents. Advanced maths, programming, writing, chemistry, physics, etc available for those interested, even at a young age.
But the baseline education should focus on learning the absolute minimum to survive and otherwise maximize fun, creativity, and empathy.
For me, and I assume others, the act of writing is an important part of the learning process. Writing on paper even more-so than typing on a keyboard.
Writing forces me to organize my thoughts, summarize, elaborate, reference other parts of the text, iterate until the pure essence is clear, remove redundancies, and this cements the concepts in my mind.
Merely reading, or worse, hearing, or worse copy/pasting something is only the first part of learning.
It's similar with programming but I would take it even further. I never really understand complicated code until I've written tests and run the debugger and been very close to it.
An AI chat bot is a powerful tool, but if you just use it to generate assignments you won't learn much. Inevitably it will be used both well and poorly and the results will be mixed.
> Writing on paper even more-so than typing on a keyboard.
How do you get to that conclusion? I find that if I have a text editor, I can write my thoughts down and then visually put them in order or encase them in a more general concept with ease, which I couldn't do when writing on paper.
It's worth noting that they were speaking about their personal experience in that paragraph. So probably for them, the "how do you get to that conclusion" is "trial and error".
But I've noticed that many many many people report the same effect, that there's something about pen-and-paper writing that's more effective for thought-lubrication. I resisted for a long time, but now I too am a convert to this school of thought.
Similarly almost everyone notices the downside: it's easier to reorder, reorganize, cross-link, etc, those thoughts, in a text editor (to say nothing of more sophisticated software tools). Some people have systems for doing complicated things with paper that they say mitigates this downside, but I am not currently one of them.
I guess it's possible that your brain just doesn't have this pattern in it. (That is, the pattern of finding pen-and-paper more effective for getting the thoughts to flow.) I mean, for all I know, maybe the huge silent majority doesn't have this pattern.
> there's something about pen-and-paper writing that's more effective for thought-lubrication
In my personal experience, a helpful feature of pen and paper is that it is less effective than keyboards, it takes me more time and focus to write things down. Maybe this gives the rest of my brain more time to catch up and understand the things that I am writing.
Written text is also less efficient when it comes to searching. This forces me to organize my thoughts better because I know I won't be able to CTRL+F random keywords later.
Another benefit of text is search, especially across documents. Also great for spellchecking and rearranging phrases, sentences, paragraphs.
But writing enables arrows, lines, crossing-out, small-text notes, circling, variable pressure, colour, etc. Richer than ordering/indenting text. Also higher contrast.
I will go even further. The physical sensation of pen on paper ... including the texture and pressure you apply to key words, capitalization, underlines, etc ... all of it being fed back though your muscles into your brain and getting processed/stored/intertwined with those very ideas and thoughts you are putting on paper ... and the mental faculties you allocate to aligning the text against the margins, ensuring neat spacing, etc ... puts your brain in a zone better tuned to the task at hand IMO.
Some of it may be just ... overload your brain so it cannot think of anything else ..so you stay focussed for lack of choice.
I mostly came to this conclusion in math classes at uni. Handwriting my notes was far superior (in terms of knowledge retained) to using my laptop. My recall was excellent. You could argue that writing math is hard in a text editor (due to the required symbols) but I think it was deeper than that. Writing on paper requires more mental focus than smashing keys, it takes longer and that feels good when you're digesting abstract concepts.
I don't write code on paper for (probably) obvious reasons, and I tend to write essays in a text editor although I also enjoy the act of writing on paper in that situation.
Different people undoubtedly come to different conclusions on this.
Personally, I find while the computer provides powerful writing tools, it also provides powerful distractions.
Maybe you get notifications you just quickly want to check, that slack message from the boss could be urgent. Maybe you decide to just check you're using that obscure word right, or to research a detail for your writing and an hour you haven't written the number of pages you set as your goal. Or maybe sitting in your netflix-watching chair looking at your netflix-watching screen just doesn't put you in the right mindset.
My habit for college essays was to write scratch notes on paper, with lots of bullet points, and arrows for re-arranging text, to get the outline of the essay in place, but to actually put the real words together on the computer.
I remember my mother once doing the opposite: she wrote a long letter on the computer, so she could edit and re-write until she was happy with it, then printed it out and copied it by hand, for the personal touch of a hand-written letter (a peacemaking letter to a relative).
> Writing is as much about organizing your thoughts and learning how to build an argument as it is about putting words on the page.
If you cannot organize your thoughts, explain your reasoning, etc. then you’re not going to get very far leveraging an LLM. Sure, it’ll spit out a book report, but unless you can explain — in well-structured writing - what you’re looking for, you’re not going to get what you need for the vast majority of writing assignments.
Essays are often about organizing other's thoughts, from references and other source material, and your own thesis based on that.
Organizing and arguing your own thought would be a good test. I think an assisting tool could still be reasonable - choosing between different organizations. Though it's unclear how to assess for original thought - the only such "essays" I know of are PhD theses.
People said similar about log tables and slide rules. And they were right - something was lost (e.g. the sense of nearby solutions). Yet here we are.
