When I had started my PhD, I thought the writing part of a paper was just the last thing I needed to do to summarize my results, like a boring epilogue to the more exciting science. What surprised me was how much of the scientific process relied on my ability to convey my ideas to someone else. And to be able to do that you need clarity of mind, you need to have worked through multiple drafts to be able to get to your point as efficiently and clearly as possible. So many times I only discovered what I was actually researching during the writing part.
The problem with AI writing is that it makes people believe that writing is just writing and not a tool for thinking very deeply about the topic you are writing about. I am not completely opposed to AI assisted writing, but if I see AI writing, it’s telling me that the author tried to take a shortcut in their own thinking and it’s a good heuristic that the ideas will not be thought through enough to warrant my complete attention.
> When I had started my PhD, I thought the writing part of a paper was just the last thing I needed to do to summarize my results, like a boring epilogue to the more exciting science.
I personally (caveat: in math, not science) always found the writing part more pleasant and exciting than the research part. In research, I was always anxious that I won't have any idea, or that the proof idea won't work. In writing, I was relaxed that I had obtained a result so I got excited about how to write it up beautifully.
Same with reading actually. Reading a book makes you think for a long time about the subject and form your own judgments. I like audiobooks and online content too, but they don’t seem to stimulate this nearly as much.
For audiobooks, I seem to not have as much of a tendency to pause the audio and stop and think, and it’s harder to reread a passage, so it’s a more passive experience.
For online content, there’s not as much of an intention to fully engage with the material. It’s easy to skim or switch to something else.
>it’s telling me that the author tried to take a shortcut in their own thinking and it’s a good heuristic that the ideas will not be thought through enough to warrant my complete attention.
Even worse: Facing AI writing, there's a strong suspicion that any concerns or counterpoints you bring won't be actually considered, but rather copypasted into an llm session.
Indeed. I used to thoroughly help strangers with programming and automation when they asked, or to explain the issues with a PR that even though I wouldn’t merge I wanted it to be clear where it could improve. It’s rewarding to teach someone who wants to learn. Nowadays there’s a lot of “I asked Claude and it said this”, so I don’t bother. I’ll gladly give you my time if you want to understand things, but I’m not going to waste it writing a thorough response just so you ignore it and shove it into an LLM.
I think you misunderstood. I believe the parent comment is talking about an audience who does care enough to message the author, but the author would just shove that message into an LLM to reply.
The trick, I think - at least this has helped me - is not to use the AI to write for me, nor to use it to rewrite my drafts, but is to do the writing myself and to ask it for editorial feedback. This forces me to think through what I’ve written more deeply, and then rewrite it according to that deeper understanding.
I think if I used AI for the actual writing then, (a) there’d be a lot missing, and (b) I’d rob myself of much of the value, in better understanding, of writing myself.
We’re not all the same so YMMV, but this is what I’ve found from using AI in various ways to see if and how it can help with writing.
With respect to seeking feedback, my angle is that: even if you do agree with the feedback then what they do. Are you going to take it as is? Are you going to come up with it and phrase it in your own way?
What next? Next, what if there's a bunch more feedback that you agree with and you incorporate it. Then, where do we go from there?
Can you even go back to (b) and make your own edits at this point? I'm not sure and I worry about a slippery slope where you stray too far from your own original form -- so I landed on not seeking feedback either.
note: this all is definitely under progress and as I'm looking at it it's very noisy with all of the formatting... just finally getting my site put together
For me I take the view that it’s similar to asking a human for feedback. I sift it, ditch what I’m not interested in or don’t agree with, do more research on what might be interesting, then decide what I want to incorporate and rewrite from there.
On grammar, typos, and the like I tend to take LLM feedback as read in most cases. Beyond this I think it’s a mistake not to do the legwork yourself, just as you’d have to if you’ve ever worked with a human editor.
But the LLM is helpful because it reduces the cycle time for feedback, which helps maintain momentum - helps me anyway.
I generally don't think LLMs writing prose for humans is a good idea (except for perhaps the most boring kinds of copywriting) but I feel a lot more positive about LLMs as reviewers, even if I don't necessarily use them myself often. My positivity comes from how relatively easy it generally is to validate feedback on a piece of writing you're intimately familiar with.
Gotta say, I don't relate to your concerns about "stray[ing] too far" _in the short term_. I think if reading LLM feedback has upset your original thinking that catastrophically you may have either been wrong (in which case, good that you received feedback early?) or writing with little intention.
However in the long term I can totally see how you could get into a habit of laziness, trusting the LLM to fix your stuff, or mindlessly applying fixes without really considering them.
One of the things I ask AI to help edit my writing is a consistency pass, making sure my ideas hang together well. The real test is to see if other humans understand it.
I tell mine that line editing and formatting are the very last polish.
I feel that way about coding. Sometimes you have an idea and a plan to implement it. But once you start coding you realize that some things were not fully right or way more complex that it should
If you let AI code for you, you will lose that
Yes. The code I produce in a year I could type in a week. The time goes into understanding the problem. You're not going to get significantly more code without reducing the level of understanding.
This happens a lot, which is why agile-style methods kept popping up to replace waterfall techniques, even before anyone tagged them with the ‘agile’ label.
Ironically, that led to a Agile™ consultant-industrial complex that tried to make it a cookie-cutter process again, but we still haven’t escaped the need to break out the IDE and actually put idea into something concrete, and I doubt we ever will.
I think that is a problem between keyboard and chair situation. I use codex in addition to my own thinking through use cases and talking to stakeholders & domain experts to augment my understanding and find gaps in my plan. In other words, I work how I worked before but with a bunch of agents helping me research and diagram out plans faster before I start coding something.
I agree, it sounds like you have a problem between your keyboard and chair, if you're generating any volume of output and not getting it in detail.
The act of vetting it and finding the failures will increase understanding, but at a cost in speed over doing the research from scratch. (Generating subtly wrong text is often a good way to test understanding, and the output of AI systems is typically a mangled clone of some other well understood system. A very good exercise is building your internal pattern recognition so you can tell what's been mangled, mashed, and served up)
AI is very good at making the mash sound plausible, however, so if you don't look extremely closely, you end up repeating instead of dissecting it. And if you're hoping to learn from the AI how to design a system, you're in trouble.
you can dump your thoughts and constraints you can think of, and have it work through untangling, or use it as a summarizing notepad, or yo prototype what you are thinking about in different variations.
its no replacement for pen and paper, but no typed or drawn on tablet picture comes close to a real notebook to scribble on
its entirely a thinking tool, as much as a skip-thinking tool
There's a difference between asking a model to figure out your argument and using it to make an argument you've already spent days working through more readable
Absolutely. And ironically one thing I discovered while trying to use AI is that it applies to programming too.
I'd heard that one-shot a project with AI, you need a fairly complete design document instead of a short prompt, so I tried to write one.
In doing that, I'd include the unhappy paths as well, e.g. in describing a list/detail interface: the list shows the X's (I thought: but in what order?) in alphabetical order; when an X is selected, the detail is displayed in the detail pane; (lists should work with multiple selection, so) multiple items may be selected at once, in which case the detail pane should display "N items selected" instead of the interface to edit the items; (a list starts empty, so) if no items are selected, the detail pane should display "No items selected", etc.
Laying out the entire system in plain English helped me have the whole thing inside my head, and see cases where presumptions about how the system would work wouldn't actually hold in practice because they weren't even holding in the text I was writing. So I had to edit the draft to fix my wrong assumptions until it made sense.
And of course the AI managed to do a pretty terrible and disappointing job at it in the end, because it made mistakes in the parts I didn't specify, or placed widgets in the wrong places because the layout that was difficult to explain in text, etc. It's somewhat impressive that anything ambiguous is a spot where AI can make the worst assumption imaginable, like it doesn't have the capacity to intuit "oh, we are making THIS sort of app, so it should look like this". But it was a good learning experience.
I suppose it's a great idea to write out what you're about to do before you set out to do it, because it helps you think about it in whole and in detail. It would be terrible if people stopped doing that.
this lands. and i find it true for software development too. a lot of the initial codebase is me discovering what i am *actually* building (should be building). all that vanishes with vibe coding.
I am starting to believe that if something is worth sharing, it is worth writing (atleast once) by a human completely. An AI clean/pass-up after that might be okay but atleast once, an author should try to create the whole piece.
I used to have a real complicated dilemma on which content exactly to write (& then share), I recently started using handy's TTS and storing it in my scratchpad app within my native language or English and then I paste it into Chatgpt and use its memories to connect with other previous discussions I had.
It helps in saving thoughts which would've otherwise been lost, even provide some new perspectives and get them in a way that I can access them later in more depth.
I think that if I were to now create blogs of these contents. I used to think that it would perhaps be okay to share an blog like thing created by AI if say I spent 10-20 minutes thinking of the rough draft of it, but after reading your comment, I am reflecting on the fact that, atleast once, people should write things themselves no matter which AI thing they use before creating blogs.
As the point of blogs shouldn't be to share unlimited content but rather that it is the specific content where a human effort was placed. Maybe if one still wants, there could be some other blog post which could be written with all the AI blogs content I have created perhaps but still, it should be seperated away from human content.
And there is this fact that as you mention, writing itself forces you to think through ideas and thoughts which are highly relevant. So previously when I used to think that re-writing them would be a waste of time and that AI writes them good enough. I think that writing them still might make a lot of sense as it would be a net benefit if not for anybody else, then for me as it can help me better understand the material myself (even more so than reading/speaking it)
The most typing/writing of myself now occurs in Hackernews comments for me and I do observe that I learn a lot writing them personally.
This happened to me recently. I just asked what the prompt was. I found it really tough to read AI words, not knowing the weight to put in each word. A almost gave it more credit than was due and responded in detail, before seeing the prompt and responding barely at all - equivalent to the thought put in.
Writing is one of the things I have chosen in my life to leave completely free of AI. It feels so important to processing information for not only myself, but for others. I often feel like I never really understand something until I write about it.
Writing code (Programming) is a thinking tool too.
When you are writing code manually, actually you are thinking about what you are doing, what you will build actually, what are your or users' real requirements, etc.
This will sound extremely stupid, but I have a skill that I use after letting an agent write code for a new feature. This skill creates a copy of the base branch and shows me each of the diffs that matter (outside of imports and other non-important boilerplate) and then I type it on an editor.
I don't use it too often because of the time it takes, but for critical work it is really good.
I have tried to come up with other equivalent ideas, like asking questions about a specific aspect of a feature, or showing the diff with some gaps to be filled, and even though they help I am still on the look out of methods to "imprint" the understanding of features into our brain.
Does anyone know any good papers or studies that could be useful in this research?
That’s so interesting. I remember trying to copy proofs in school out by hand but it did not work for me because - I think - I didn’t engage with the material as much as I should’ve. How do you ensure that you’re not simply blindly typing the code out and actually understand what’s going on?
Well, I only do it in order to try to understand it but there is no way to make sure you are understanding. But most of the time, if I am typing it's almost automatic that I pay attention to it. Not sure if it would work for others tho.
Walter J. Ong wrote an essay a good long awhile ago entitled "Writing is a Technology that Restructures Thought"[1] which is worth reading on this topic.
I think where this technology is heading, AI will do all that much faster and better. Humans are still under the illusion that they need to be in control.
This is true for designing, too. I see a lot of people saying Figma is dead because you can prototype directly; for novel UX problems the act of designing — of putting ideas together, assembling a component, even getting good data into a design — is where a lot of my thinking and insights develop. The brain work plus the craft work, the tight loop between thinking and creating is so critical for me and I haven’t found nearly the same feeling when prompting.
In writing and designing I think we often have a thought-feeling first: something inchoate that we can’t quite articulate yet. Getting directly to the medium itself — text or frames and layouts — is how we give shape to that feeling and turn it into a concept, insight or design.
I see what you are saying and for very specific problems/explorations I do still reach for Figma. But the days of crafting static screens and multi-artboard flows are gone. It's better to prototype the entire flow/interaction directly than trying to reason about it 'statically'. LLMs have also opened up an entire array of new possibilities for e.g. combining meaningful 3D and taking a 'leap of faith' into visualizations/interactions that would simply be too costly to properly explore before
> But the days of crafting static screens and multi-artboard flows are gone. It's better to prototype the entire flow/interaction directly than trying to reason about it 'statically'.
I haven’t found this to be true but it could well come down to the type of design work I do. I’m usually working on new features for complex workflows within an already complex app. When I’ve tried to yolo designs with LLMs they never come up with elegant visual or UX solutions. They’ve been occasionally useful as a way to get something on the screen that I can work against — ie “well that’s clearly wrong in these 20 ways” — but at that point I usually then reach for figma so I can just do the damn thing myself.
Agree about the more complex design options it opens up, I just haven’t had call to do much in those spaces yet.
I actually appreciate working with LLMs on design quite a lot, although I'm not a designer by trade so my situation may differ. I enjoy getting an image-generated mockup (that is usually top class), then letting an LLM implement it, then seeing where it doesn't feel quite right. Then trying to understand why it doesn't feel right. When I do the CSS myself, I iterate one component at a time and usually not get the 'find the 100 differences' feel. Animations are even more fun, you feel they suck, but you have to slow them down to understand what's really making them suck. Not claiming I can do good designs, but it helped me identify faster why some designs were bad.
Yeah I can see when you’re on a certain part of a learning curve how that kind of iteration can be really useful. “That seems good, let’s see it. Hm now I feel how this part is clunky”. I can imagine looping that way for certain kinds of things I don’t have a ton of experience with — animations will be a good one for me actually.
Most of my projects are born out of scribbles on paper, system diagrams with boxes and little arrows to try to describe the system I want to build or my thought process written down like in this post.
When I get it on paper and everything feels right I start creating the codebase but I have to get it down first, feels like I remember it more after I save my thoughts to external paper storage.
I have over the decades tried every combination I can think up or have read about for designing things and I always land back in the same place, A4 5mm square pad (cheap but not the cheapest) and a really good (but not expensive) set of ballpoints in black/blue/red/green (and mostly it's black anyway).
There is something about the complete free form nature of pen and paper that makes me think better, I bought a nice ipad/pencil a few years ago thinking that would replicate it but it didn't, it was closer but not close.
If it's something I want to keep I just take a photo but vast majority of notes are transient anyway.
I rediscovered my university revision notes recently and I don't think it's a coincidence that the year I got my best grade was the year I went to town on very neat, carefully laid out lecture notes. Whether I was taking exquisite care over circuit diagrams or deciding which colour to make streamlines and velocity gradients, it helped bury the facts in my brain.
I spent an unreasonable amount of time recently thinking about how to color- (and symbol-) code my notes. I wanted to create a simple generally applicable system that I can use for any project.
For thinking, nothing beats handwriting. It's just the unconstrained freedom of movement in every direction, shape or form, and the lack of friction and distractions.
But then you sometime also want to organize or share stuff, so that's where digital comes in. I have an automation that takes a photo of my notes with my phone, uploads it to Obsidian, and OCRs it into markdown and tags it, so I basically end up with a "symbolic" 1-1 equivalent of analog handwriting + digital text
My system still evolves, but what I currently use is:
Color:
- black = just "normal" text
- red = critical/"must"
- yellow = important/"should"
- green = highlighted/"could"
Symbols:
- bulletpoint
[ ] task
-> reminder
I also do brackets and underlines which translate to # heading1 and ## heading2 in markdown
I write by hand with a fountain pen with black ink + use 3 fineliners and small colored sticky flags.
Some observations:
- yellow is not great on white paper due to bad contrast; it also rarely comes as ballpoint, but fineliner options are available
- blue and green are visually very similar, to the point of kinda being redundant
- for sharing with other people, the system must be as generally understandable as possible. Here the traffic norms such as ISO 7010 (https://en.wikipedia.org/wiki/ISO_7010) could be considered
- also there's the aspect of accessibility, with a relevant portion of population being colorblind
- The simplest most effective system seems to be: one main color (black or blue) and one highlight color (red I think works best), plus a couple of symbols such as bullet point, task, reminder that translate well into markdown
I use the "uni-ball - Jetstream SXN-210" (in 1mm), they are superb, don't smear, write smooth every time and are cheap enough if I lend one to someone and don't get it back I don't regret it, ergonomically they are comfortable as well, not a lot else to say about them beyond that - just a really well designed pen.
Agreed, I used to role my eyes when my professors asked me to do things like pseudocode (which I still don't do too much) but the act of writing an outline for the general structure of a program always helped me get over the initial hump of starting a new project.
For me, writing is also a thinking tool, that is why I don't use AI to write. But using AI to generate ideas for writing? All the time :D (That way, I don't need to wait for that inspiration). I like the idea of scheduling it though!
One of the biggest thing we lose as we replace workflows with AI are the ancillary benefits of the process of writing (prose or code), design, or anything else.
You may produce an acceptable result faster with AI, but the deep understanding, questioning your own assumptions, and correcting your own errors only come from the actual writing process.
And these are not nice to haves because that cognitive context drives all downstream work.
When i started writing i felt the pressure to always go from draft to a polished piece. i posted a couple things i worked semi hard on and received 0 views/reads because i don't really have much of an online presence at all.
From there i kinda just moved on and write purely for myself, and use it to up my own thoughts. Very rarely, something will receive a second pass days later, and thats just more to go back on my past thoughts and try to tighten them if i feel that is needed.
I tend to split my day into two parts, my work hours and my "unplug" hours, and the writing was mostly landing in those unplug hours as a way to sorta summarize my day. Last week or so I decided im moving that writing into the very first thing I do each day, and it has really been absurdly effective at getting new ideas flowing quickly. I used to depend on my later day brain that was somewhat "depleted", for lack of a better way to describe how i work, and from there that would kinda birth what i work on the next morning, with maybe some iteration while working.
Now instead I just do both, I open the day with ~500-1500 words, just anything that comes out, then i potentially get stuck, and ill maybe dump that to a model as just a modern rubber duck, not anything like "write for me, or finish this thought", just a kinda nudge to really toss back at me what im fully considering and then maybe i pivot back and chase down the most interesting of the topics for a few hundred more words, or get straight to work on what i thought was the most vital of all those ideas to chase while im fresh in the morning.
But, to the claim of how to think in the age of LLMs, I think it’s worth contrasting the other approach, which is a dialog with the LLM; two things are needed IMO to keep things honest:
First, you need to keep asking the LLM to explain things, until you understand what it did. Lots of angles here; have the LLM make a design/product doc, write one-off architecture explainers, write throwaway data-viz tools to explain the data distribution. Then ask a fresh copy to adversarially review the core concepts. Try asking a different frontier model too.
Second, you need to push for “proof”, in other words, ask for demos and try it yourself. Imagine yourself a CEO or VC; how do you know if the output of an entire org is vaporware? One thing to beware is launching into a field where you don’t have expertise and can’t actually validate the output. This takes time of course, but for example if you write a stress test script and get CPU/QPS numbers, these complex validators are IME quite reliable when you apply the same rigor from 1) above to their construction.
I'd been a software engineer for 25 years before language models came into the scene. There were so many parts of the engineering world I had no idea about: Large cloud data processing, lower level device / GPUs, UI frameworks and design.
I kept my to my niche.
Now I can build anything, but as I'm building it, I try to understand at least how it works. So I'm constantly asking questions. In the process, I think of my prior engineering skill as somewhat childlike. As I thought I was a decent engineer, I now see that I was decent at some things, but far from the kind of breadth that the wide engineering world has.
I think the takeaway is that you should always be curious and build things. Whether it's writing or discussing with an language model, writing software by hand, or creating software with a model and understanding it, I think you have the same effect.
> Now I can build anything, but as I'm building it, I try to understand at least how it works. So I'm constantly asking questions.
This is the way!
AI is the best technology ever invented for avoiding learning things. It’s also the best technology ever invented for learning things.
I think it’s very important to be on your guard for accidentally glazing over your understanding, but if understanding is your goal, there are so many more tools for this now.
Before AI, even when working collaboratively and bumping into concepts I did not fully grasp, there was so much more friction to get answers. Sure I could Google and try my luck, or I could ask a colleague. But the LLMs have infinite patience, can build diagrams, and can answer weird questions I would be afraid to ask.
Words are a communicable but ultimately lossy map of reality, experience of which we cannot otherwise convey; if the goal, the final artefact is itself a map (e.g., a thesis), using writing at earlier stages is, in a way, beginning to sketch out said map.
I endlessly enjoyed the comparison with a Turning Machine. When I study someone's complex work, I do notes to keep several things in my sight scope: a piece of paper, PowerPoint slide or some fancy markdown text editor with syntax highlight. Seems like that this focusing is somewhat the natural way of building neuronal connections: if a logical piece is missing we have to use imagination to complete it and only then it nicely fits into our brain's "hologram" of the world.
There is a wonderful book "Exploring Writing" by John Langan. I found it in our library when I was in my PhD writing phase. I highly recommend it to anyone who has problems building text and who wants to learn how to master its structuring.
And for freewriting, sometimes when it gets really hard to move further: https://3am.freewriter.app/
If you set 5 min target duration, you will be awarded with a full version of the song!
I always carry a small Moleskine notebook, the pocket-sized hardcover with grid paper, for notes, thoughts, and whatever else comes to mind during the day. Each page is numbered, and I continue the numbering as I move from one notebook to the next. This allows quick reference to past ideas, thoughts, or notes I may have in an older notebook, or even a few pages back. It is the primary way I get thoughts to paper, and is the "entry point" to all my writing and thinking.
Notebooks do not crash, do not freeze, and need only you and a pen. My notebooks often remind me of this passage from "Zen and the Art of Motorcycle Maintenance":
> “For this you keep a lab notebook. Everything gets written down, formally, so that you know at all times where you are, where you’ve been, where you’re going and where you want to get. In scientific work and electronics technology this is necessary because otherwise the problems get so complex you get lost in them and confused and forget what you know and what you don’t know and have to give up.”
> Robert M. Pirsig, Zen and the Art of Motorcycle Maintenance
It's an attractive thought, this comparative look at FA and TM. However it is misleading to an extent.
A Turing Machine's capabilities exceed the FA's because it has a 'store of knowledge'. It is both recording and reading, repeatedly. There is erasure, too. So if we're aspiring to TM-hood, the guidance seems to be write, and re-read (your own thoughts), and re-evaluate, and re-write. This process is subject to the halting problem. /g
An observation as it relates to LLMs, anecdotal. I noticed that when I screenshot a handwritten scratchpad of my thinking, rather than just write a prompt with same thinking, they seem to be more “eager” to respond and more keen to understand the ideas. Maybe an illusion but if anyone else noticed this it would make two of us.
Writing for me anyway is more like a general anxiety release valve.
Thinking is we just do without external tools as we reassess frame to frame due to the world around us not being static; we develop object permanence of a tree for a while but are burning calories to replace on a much shorter timeline; whole lot of physical state still updating on shorter timelines than a tree; so our brains automatically always wipe and refresh the canvas as a natural consequence of relative entropy
We track these deltas and how to deal with them; writing helps deal with scenarios where there is no roof to fix to stay sheltered. What do I do with these thoughts with nowhere to go; I have food, I have shelter? Writing.
Reading Murat's blog around the start of the year introduced me to the idea of freewriting. I built a tool that I still use weekly for coming up with new ideas / exploring thoughts.
I have seen the same thing happen in multiple domains. You may have identified a problem to solve, but the act of solving it, by writing it down, by writing a program, building a tool, making a device, or even carving from stone.
As you work on it the problem becomes more clear and seems to tell you what the actual shape of the problem is and you have to adapt your solution to that actual problem, even if it ignores certain aspects of the original idea.
Writing it down and trying to clarify it is one of the cheapest ways to get to the real answer.
A common refrain for teachers and writers is that they learned so much producing their work. You think you know it, and then you go to teach it and realize how much you didn't know.
I'm a big fan of marbled composition books. They're cheap and durable enough. You can write in them without a table. I have 5-6 of them going at once for various tasks. And somehow (for me) the ideas flow more freely longhand than they do at the keyboard. When full, you can cut the spine off with a box cutter and scan it. G4 compression loses the color, but makes really small PDFs.
It's easy to produce better written content than an LLM. And that's good for the reader and the writer.
I have been finding I spend more time writing when working with AI. While I do let it generate docs, it does not always do a great job and I end up reorganizing, reframing, or changing vocabulary. Those passes end up require rigorous thinking.
There has been some occasions where I cannot use AI as a thinking partner. It doesn’t have the imagination, intuition, or judgement. I have been relying more on my team in those cases.
This article reminds me of another method: grabbing paper and pencil and start doodling around to work out some ideas.
There's always this peculiarity regarding the fact that writing by hand, is not the same as writing by kbd. So I'm always confused which of these two actually authors, mean, though I agree - any sort of note taking is better than none.
For me I think the distinction is more about structure and rigidity than keyboard vs pencil. Sketching free-form notes and especially little diagrams is a really powerful thinking tool. I also like to sketch my thoughts in org mode, which has some advantages over paper (structural editing, ability to zoom in and out by folding subtrees) and some disadvantages (tree structure instead of arbitrary spatial structure, more friction for diagrams).
I have a really hard time using WYSIWYG editors for thinking-by-writing. I think it’s because it’s hard to see partially-formed structure so it feels less malleable.
Yes, me too! I like using sketch notes to think through things nonlinearly, then I make an outline in Org or use an AI to get the text from a sketch, and then maybe I'll refine the sketch so it can be an illustration that
I embed back into Org. Structural editing is fantastic for reordering thoughts and converting back and forth between lists and paragraphs/sentences.
but then, from personal experience, once you master, and practice photoshop/illustrator-like software, it becomes not so hard to actually produce quite complex visual communication in little to no effort. it is questionable whether I'd ever be able to do so with pen and paper, though, of course, the feeling... the touch of it is second to none. Talking of practicality though, it makes me wonder. I think rather an important question is whether we also talk about writing/note taking as part of larger personal information management system. It is very likely having one (such own KB) matters even more than the medium itself.
You can get the best of both worlds by picking up a Wacom tablet or similar. It isn't exactly the same, but it does then mean you can doodle but with the power of the capable software.
definitely thinking about it every day, as the pain in my wrist, one resulting from touch and mouse use, is increasing with no end in sight. some pencil and paper based loop through the LLMs seems even more appealing.
Writing is not helpful just because it has Turing-tape memory or because it reduces noise in your signal to others.
It's mainly helpful in changing an exercise for youself into something useful for others, that requires you actually adopt their perspectives and migrate them to yours (or realize the error of your ways). You take on the capability not just to convey ideas but also to change the world.
Interestingly what has worked very well for me is that when I start a new project, I just have to start drawing a few boxes and connect them as I expect them to be communicating and with just an hour or so of doing this, I find myself ready to write code as if there was a block that got completely untangled with putting it on paper.
> Yet the only difference between a FA and a TM is that
> the TM, unlike the FA, has paper and pencil.
> Think about it.
The analogy doesn't works if you really think about it. A TM with a finite tape is equivalent to an FA. The fundamental difference between a TM and an FA is that a TM has infinite space.
(Yes, I know I missed the point and am not very fun at parties.)
No, this is a very good point - you need a supply of pen and paper, such that it feels effectively infinite. Buying a nice notebook or a nice pen is a trap because both feel finite; buying a box of decent pens and a carton of decent notepads takes you out of the scarcity mindset and unlocks the Turing machine like advantages mentioned.
I was thinking of the same. FSA can multiply within bounds. For FSA, the storage is the state itself and bounded.
I agree with the sentiment but not a fan of analogies like this. We do have memory even if we don't write. Each organ adds to the memory imprint making it stronger. Visual, smell, taste, audio etc Each action adds more imprints. And this makes more sense to me than an analogy.
Another analogy which never made sense to me was improving by just 1% every day (1.01³⁶⁵).
Analogies do help explain phenomena I cannot even relate to at all, like telling me how life on Mars would be.
Ofcourse I'm not invited to parties as you can imagine, taking things literally.
A big part of understanding something is getting used to it, and analogies help with getting over a mental hang up. They may also help with deciding which areas of the new space to explore first.
But we should be examining the limits of applicability, as we do for most cognitive tools, so we know when to switch from one tool to another. It also helps to combine such tools in a complementary manner.
Are you saying that 1.01^365 is too big? I think there are lots of things where if you are starting from near zero skill, that's not a crazy big number.
Not parent commentor, but My issue with 1.01^365 analogy is that it implies that somehow that skill multiplies itself. There is no obvious reason to believe the gain from _(skill acquisition, language learning, workout, whatever) is better modeled as 1.01^365 than 1+(0.1*365).
Being forced to take notes and write essays in school made me hate handwriting specifically and dislike writing in general. Now that I have a choice, I never use stationery or take notes for any purpose, and everything useful that I learned or did in my entire life has been done without writing being a part of the process.
> Don't be discouraged by how bad the first draft was. McPhee says the way to do a piece of writing is three or four times over, never once. This is good news.
This also applies to code. Taking your agent's first draft as a sample and doing a reimplementation often yields an even cleaner version.
Once you do ship, the drafts' weaknesses provide codebase signal. Where should you shore up the architecture - or create more reusability, where is the documentation inadequate. Do you need to bake invariants deeper in so they're harder to violate (custom lints, stricter validation...)
The human mind has a small L1 cache. High IQ people tends to have larger L1 cache, and don't understand when others can't follow their arguments. Not understanding that's it's dumb.
Writing as a thinking tool has become sort of a meme these days.
And, yes, I also generally believe it’s a good thing to use it as a tool. I try to do use it as often as I can.
But… I think it’s also important to note that there are - like with almost anything - pitfalls and downsides. I rarely see them discussed and thus want to bring them up:
For highly analytical people, writing may become counterproductive. Especially, when it is used to continuously elaborate thoughts rather than to challenge assumptions, test hypotheses or reach decisions.
It can lead to over-analyzing or pseudo-rationalizing.
Beyond all this, there's also tailoring your message to your audience.
In sales, you might call this finding "words that work" or "the pain points". AI will not know how to do this because it's getting a vibe check of the room.
You cannot merely be a sensory part for the AI unless you want to sound as cringe and out-of-touch as the majority of the training data (stale social media). Garbage in, garbage out.
I spent a lot of time thinking about it. AI converges on average which is very bland. The labs are trying to get out of it with RLFH which sadly just produces a different “edgy” average (Calude-speak).
But things get better if you instruct AI to speak in certain style. It’s still average but for the chosen style and thus less tone-deaf for a given use case.
So my thinking is - how much mileage can we get by introducing specific style guides? We can create a library of styles, and to get out of average for a specific guide individual companies and even people will maintain their own private patches from upstream.
Another option is to maintain history of your own writing and thus preserve your own voice. I feel like “averaging” of style is somehow un-ergonomic, in that in numbs the mind, and thus even diversity for the sake of diversity is valuable.
Words are most effective when they buck the trend in just the right way.
I'll use the conference room example for clarity, even though a lot of people still work remote these days.
Reading the room could be as subtle yet significant as reacting to the speaker that was on immediately before you. It could be reacting to the temperature of the room or the catered lunch everyone just had. It can be noticing or participating in or critiquing a peculiarity of the audience (hawaiian shirt fridays, etc.). It can also be a million other things and they'll never let you get away with choosing at random. You have to just "get it" or you're going to be ignored at best.
You're at once trying to both signal in-group status while also asserting your own identity and setting the tone for everything you'll be saying after. Remote work has its own idiosyncrasies, but regardless, you need a pulse for this kind of work. And, all this is just the introduction. What about emphasis choices in the rest of your writing/speaking?
Trying to pin that down is simply not logical and it's not just a matter of "adding" a handful of tricks unless you want to alienate everyone. Sometimes it's also the things you don't say or do that make the difference. You also have to accept that you can't please everyone. You could actually piss off everyone but the most important person in the room and still win. That might actually be the best way or the only way to win sometimes (as was the case for the American sales pitch of AI itself!).
The real world is messy and adding AI, or any particular process, to the mix just makes it worse. You cannot plan an accident, but you can avoid one.
I drove my teachers nuts with my unconventional writing process. I start with a blank word processor document, and start making notes. Then I rearrange them, and finally I start turning it into an essay. My teachers stood over me, demanding that I stop misbehaving, and write an essay from start to finish. I could not do that. My brain does not work that way. I can not keep enough things in my head to organize a few pages that way.
---
I was once thrown out of a job interview after their secretary yelled at me for what seemed like 5 minutes because I was filling out a form from the shortest blank to the longest blank rather than from the top to the bottom. I had experienced such a trying day, that I just could not bring myself to fill out anything but the next shortest field. All the while, this crazy lady was yelling at me as I had to stand at her desk in front of a room of HR scum/manager scum and job applicants.
---
I classify linear thinking as one useful way to check if something makes sense, not the only way to think.
I sense a hint of malicious compliance. Did you also write all the vertical strokes, then the horizontal ones, then the loops? Vowels, then consonants? Write each letter in alphabetical order? Express your zipcode as a geometric series? You have a future in the go-slow bureaucracy of your choice! :-)
"Complexity in prose is usually a place to hide sloppy thinking, or a way to weasel out of taking a stance. If you can't say it plainly, you probably haven't thought it through yet."
The place this surprised me most was before writing code. On a recent project I wrote a plain plan file first: numbered stages, what each one should do, what "done" means. Half the real decisions happened while writing it, not while building. Questions like "what happens when the external service is down" only showed up because I had to put the step into a sentence.
It also made the work resumable. Coming back after a few days, the file told me exactly where I was and why things were the way they were, which no amount of reading my own code did.
When I had started my PhD, I thought the writing part of a paper was just the last thing I needed to do to summarize my results, like a boring epilogue to the more exciting science. What surprised me was how much of the scientific process relied on my ability to convey my ideas to someone else. And to be able to do that you need clarity of mind, you need to have worked through multiple drafts to be able to get to your point as efficiently and clearly as possible. So many times I only discovered what I was actually researching during the writing part.
The problem with AI writing is that it makes people believe that writing is just writing and not a tool for thinking very deeply about the topic you are writing about. I am not completely opposed to AI assisted writing, but if I see AI writing, it’s telling me that the author tried to take a shortcut in their own thinking and it’s a good heuristic that the ideas will not be thought through enough to warrant my complete attention.
> When I had started my PhD, I thought the writing part of a paper was just the last thing I needed to do to summarize my results, like a boring epilogue to the more exciting science.
I personally (caveat: in math, not science) always found the writing part more pleasant and exciting than the research part. In research, I was always anxious that I won't have any idea, or that the proof idea won't work. In writing, I was relaxed that I had obtained a result so I got excited about how to write it up beautifully.
Same with reading actually. Reading a book makes you think for a long time about the subject and form your own judgments. I like audiobooks and online content too, but they don’t seem to stimulate this nearly as much.
For audiobooks, I seem to not have as much of a tendency to pause the audio and stop and think, and it’s harder to reread a passage, so it’s a more passive experience.
For online content, there’s not as much of an intention to fully engage with the material. It’s easy to skim or switch to something else.
> For online content, there’s not as much of an intention to fully engage with the material. It’s easy to skim or switch to something else.
Thanks for writing that down. It explains in simple terms the difference
>it’s telling me that the author tried to take a shortcut in their own thinking and it’s a good heuristic that the ideas will not be thought through enough to warrant my complete attention.
Even worse: Facing AI writing, there's a strong suspicion that any concerns or counterpoints you bring won't be actually considered, but rather copypasted into an llm session.
Indeed. I used to thoroughly help strangers with programming and automation when they asked, or to explain the issues with a PR that even though I wouldn’t merge I wanted it to be clear where it could improve. It’s rewarding to teach someone who wants to learn. Nowadays there’s a lot of “I asked Claude and it said this”, so I don’t bother. I’ll gladly give you my time if you want to understand things, but I’m not going to waste it writing a thorough response just so you ignore it and shove it into an LLM.
That's a good thing. The writer who has nothing to say and the dumb audience who doesn't care stay down where they belong pacified by their "AI".
I think you misunderstood. I believe the parent comment is talking about an audience who does care enough to message the author, but the author would just shove that message into an LLM to reply.
The trick, I think - at least this has helped me - is not to use the AI to write for me, nor to use it to rewrite my drafts, but is to do the writing myself and to ask it for editorial feedback. This forces me to think through what I’ve written more deeply, and then rewrite it according to that deeper understanding.
I think if I used AI for the actual writing then, (a) there’d be a lot missing, and (b) I’d rob myself of much of the value, in better understanding, of writing myself.
We’re not all the same so YMMV, but this is what I’ve found from using AI in various ways to see if and how it can help with writing.
This is my stance, but I go further to say that that you shouldn't even use it for feedback...
The feedback bit is surely more debatable, but I would argue the initial writing and allowing edits are not.
https://antoaenono.com/write/fragments/writing#:~:text=never...
With respect to seeking feedback, my angle is that: even if you do agree with the feedback then what they do. Are you going to take it as is? Are you going to come up with it and phrase it in your own way?
What next? Next, what if there's a bunch more feedback that you agree with and you incorporate it. Then, where do we go from there?
Can you even go back to (b) and make your own edits at this point? I'm not sure and I worry about a slippery slope where you stray too far from your own original form -- so I landed on not seeking feedback either.
note: this all is definitely under progress and as I'm looking at it it's very noisy with all of the formatting... just finally getting my site put together
also not sure about my fish progress bar hah
For me I take the view that it’s similar to asking a human for feedback. I sift it, ditch what I’m not interested in or don’t agree with, do more research on what might be interesting, then decide what I want to incorporate and rewrite from there.
On grammar, typos, and the like I tend to take LLM feedback as read in most cases. Beyond this I think it’s a mistake not to do the legwork yourself, just as you’d have to if you’ve ever worked with a human editor.
But the LLM is helpful because it reduces the cycle time for feedback, which helps maintain momentum - helps me anyway.
> For me I take the view that it’s similar to asking a human for feedback
But it's not asking a human for feedback. That's kind of an important difference, isn't it?
No human tried to understand your writing, and tried to explain their understandings back to you. How much is left of the editorial process then?
I think you make an interesting point.
I generally don't think LLMs writing prose for humans is a good idea (except for perhaps the most boring kinds of copywriting) but I feel a lot more positive about LLMs as reviewers, even if I don't necessarily use them myself often. My positivity comes from how relatively easy it generally is to validate feedback on a piece of writing you're intimately familiar with.
Gotta say, I don't relate to your concerns about "stray[ing] too far" _in the short term_. I think if reading LLM feedback has upset your original thinking that catastrophically you may have either been wrong (in which case, good that you received feedback early?) or writing with little intention.
However in the long term I can totally see how you could get into a habit of laziness, trusting the LLM to fix your stuff, or mindlessly applying fixes without really considering them.
One of the things I ask AI to help edit my writing is a consistency pass, making sure my ideas hang together well. The real test is to see if other humans understand it.
I tell mine that line editing and formatting are the very last polish.
I feel that way about coding. Sometimes you have an idea and a plan to implement it. But once you start coding you realize that some things were not fully right or way more complex that it should If you let AI code for you, you will lose that
Yes. The code I produce in a year I could type in a week. The time goes into understanding the problem. You're not going to get significantly more code without reducing the level of understanding.
"Everybody has a plan until they get punched in the mouth (at runtime)." - Mike Tyson
I wrote this in another comment too, but one thing that helps in this case is to write what you want to implement in English first.
If it doesn't make sense in English, it won't make sense in code. But English is easier to draft and edit than code.
This happens a lot, which is why agile-style methods kept popping up to replace waterfall techniques, even before anyone tagged them with the ‘agile’ label.
Ironically, that led to a Agile™ consultant-industrial complex that tried to make it a cookie-cutter process again, but we still haven’t escaped the need to break out the IDE and actually put idea into something concrete, and I doubt we ever will.
The problem with AI is that it's entire selling point is saving time by not needing to think very deeply about topics.
The writing is downstream of that.
I think that is a problem between keyboard and chair situation. I use codex in addition to my own thinking through use cases and talking to stakeholders & domain experts to augment my understanding and find gaps in my plan. In other words, I work how I worked before but with a bunch of agents helping me research and diagram out plans faster before I start coding something.
I agree, it sounds like you have a problem between your keyboard and chair, if you're generating any volume of output and not getting it in detail.
The act of vetting it and finding the failures will increase understanding, but at a cost in speed over doing the research from scratch. (Generating subtly wrong text is often a good way to test understanding, and the output of AI systems is typically a mangled clone of some other well understood system. A very good exercise is building your internal pattern recognition so you can tell what's been mangled, mashed, and served up)
AI is very good at making the mash sound plausible, however, so if you don't look extremely closely, you end up repeating instead of dissecting it. And if you're hoping to learn from the AI how to design a system, you're in trouble.
but in actuality, its all about how you use it.
you can dump your thoughts and constraints you can think of, and have it work through untangling, or use it as a summarizing notepad, or yo prototype what you are thinking about in different variations.
its no replacement for pen and paper, but no typed or drawn on tablet picture comes close to a real notebook to scribble on its entirely a thinking tool, as much as a skip-thinking tool
There's a difference between asking a model to figure out your argument and using it to make an argument you've already spent days working through more readable
Absolutely. And ironically one thing I discovered while trying to use AI is that it applies to programming too.
I'd heard that one-shot a project with AI, you need a fairly complete design document instead of a short prompt, so I tried to write one.
In doing that, I'd include the unhappy paths as well, e.g. in describing a list/detail interface: the list shows the X's (I thought: but in what order?) in alphabetical order; when an X is selected, the detail is displayed in the detail pane; (lists should work with multiple selection, so) multiple items may be selected at once, in which case the detail pane should display "N items selected" instead of the interface to edit the items; (a list starts empty, so) if no items are selected, the detail pane should display "No items selected", etc.
Laying out the entire system in plain English helped me have the whole thing inside my head, and see cases where presumptions about how the system would work wouldn't actually hold in practice because they weren't even holding in the text I was writing. So I had to edit the draft to fix my wrong assumptions until it made sense.
And of course the AI managed to do a pretty terrible and disappointing job at it in the end, because it made mistakes in the parts I didn't specify, or placed widgets in the wrong places because the layout that was difficult to explain in text, etc. It's somewhat impressive that anything ambiguous is a spot where AI can make the worst assumption imaginable, like it doesn't have the capacity to intuit "oh, we are making THIS sort of app, so it should look like this". But it was a good learning experience.
I suppose it's a great idea to write out what you're about to do before you set out to do it, because it helps you think about it in whole and in detail. It would be terrible if people stopped doing that.
this lands. and i find it true for software development too. a lot of the initial codebase is me discovering what i am *actually* building (should be building). all that vanishes with vibe coding.
Part of the problem is that the English word "writing" refers to both the process and the result.
I am starting to believe that if something is worth sharing, it is worth writing (atleast once) by a human completely. An AI clean/pass-up after that might be okay but atleast once, an author should try to create the whole piece.
I used to have a real complicated dilemma on which content exactly to write (& then share), I recently started using handy's TTS and storing it in my scratchpad app within my native language or English and then I paste it into Chatgpt and use its memories to connect with other previous discussions I had.
It helps in saving thoughts which would've otherwise been lost, even provide some new perspectives and get them in a way that I can access them later in more depth.
I think that if I were to now create blogs of these contents. I used to think that it would perhaps be okay to share an blog like thing created by AI if say I spent 10-20 minutes thinking of the rough draft of it, but after reading your comment, I am reflecting on the fact that, atleast once, people should write things themselves no matter which AI thing they use before creating blogs.
As the point of blogs shouldn't be to share unlimited content but rather that it is the specific content where a human effort was placed. Maybe if one still wants, there could be some other blog post which could be written with all the AI blogs content I have created perhaps but still, it should be seperated away from human content.
And there is this fact that as you mention, writing itself forces you to think through ideas and thoughts which are highly relevant. So previously when I used to think that re-writing them would be a waste of time and that AI writes them good enough. I think that writing them still might make a lot of sense as it would be a net benefit if not for anybody else, then for me as it can help me better understand the material myself (even more so than reading/speaking it)
The most typing/writing of myself now occurs in Hackernews comments for me and I do observe that I learn a lot writing them personally.
This happened to me recently. I just asked what the prompt was. I found it really tough to read AI words, not knowing the weight to put in each word. A almost gave it more credit than was due and responded in detail, before seeing the prompt and responding barely at all - equivalent to the thought put in.
Writing is one of the things I have chosen in my life to leave completely free of AI. It feels so important to processing information for not only myself, but for others. I often feel like I never really understand something until I write about it.
Writing code (Programming) is a thinking tool too.
When you are writing code manually, actually you are thinking about what you are doing, what you will build actually, what are your or users' real requirements, etc.
This will sound extremely stupid, but I have a skill that I use after letting an agent write code for a new feature. This skill creates a copy of the base branch and shows me each of the diffs that matter (outside of imports and other non-important boilerplate) and then I type it on an editor.
I don't use it too often because of the time it takes, but for critical work it is really good.
I have tried to come up with other equivalent ideas, like asking questions about a specific aspect of a feature, or showing the diff with some gaps to be filled, and even though they help I am still on the look out of methods to "imprint" the understanding of features into our brain.
Does anyone know any good papers or studies that could be useful in this research?
That’s so interesting. I remember trying to copy proofs in school out by hand but it did not work for me because - I think - I didn’t engage with the material as much as I should’ve. How do you ensure that you’re not simply blindly typing the code out and actually understand what’s going on?
Well, I only do it in order to try to understand it but there is no way to make sure you are understanding. But most of the time, if I am typing it's almost automatic that I pay attention to it. Not sure if it would work for others tho.
As long as you generate code in small increments and intentionally I don't think its a big deal.
Walter J. Ong wrote an essay a good long awhile ago entitled "Writing is a Technology that Restructures Thought"[1] which is worth reading on this topic.
[1]: https://twinada.wordpress.com/wp-content/uploads/2012/09/708...
I think where this technology is heading, AI will do all that much faster and better. Humans are still under the illusion that they need to be in control.
This is true for designing, too. I see a lot of people saying Figma is dead because you can prototype directly; for novel UX problems the act of designing — of putting ideas together, assembling a component, even getting good data into a design — is where a lot of my thinking and insights develop. The brain work plus the craft work, the tight loop between thinking and creating is so critical for me and I haven’t found nearly the same feeling when prompting.
In writing and designing I think we often have a thought-feeling first: something inchoate that we can’t quite articulate yet. Getting directly to the medium itself — text or frames and layouts — is how we give shape to that feeling and turn it into a concept, insight or design.
I see what you are saying and for very specific problems/explorations I do still reach for Figma. But the days of crafting static screens and multi-artboard flows are gone. It's better to prototype the entire flow/interaction directly than trying to reason about it 'statically'. LLMs have also opened up an entire array of new possibilities for e.g. combining meaningful 3D and taking a 'leap of faith' into visualizations/interactions that would simply be too costly to properly explore before
> But the days of crafting static screens and multi-artboard flows are gone. It's better to prototype the entire flow/interaction directly than trying to reason about it 'statically'.
I haven’t found this to be true but it could well come down to the type of design work I do. I’m usually working on new features for complex workflows within an already complex app. When I’ve tried to yolo designs with LLMs they never come up with elegant visual or UX solutions. They’ve been occasionally useful as a way to get something on the screen that I can work against — ie “well that’s clearly wrong in these 20 ways” — but at that point I usually then reach for figma so I can just do the damn thing myself.
Agree about the more complex design options it opens up, I just haven’t had call to do much in those spaces yet.
I actually appreciate working with LLMs on design quite a lot, although I'm not a designer by trade so my situation may differ. I enjoy getting an image-generated mockup (that is usually top class), then letting an LLM implement it, then seeing where it doesn't feel quite right. Then trying to understand why it doesn't feel right. When I do the CSS myself, I iterate one component at a time and usually not get the 'find the 100 differences' feel. Animations are even more fun, you feel they suck, but you have to slow them down to understand what's really making them suck. Not claiming I can do good designs, but it helped me identify faster why some designs were bad.
Yeah I can see when you’re on a certain part of a learning curve how that kind of iteration can be really useful. “That seems good, let’s see it. Hm now I feel how this part is clunky”. I can imagine looping that way for certain kinds of things I don’t have a ton of experience with — animations will be a good one for me actually.
Most of my projects are born out of scribbles on paper, system diagrams with boxes and little arrows to try to describe the system I want to build or my thought process written down like in this post.
When I get it on paper and everything feels right I start creating the codebase but I have to get it down first, feels like I remember it more after I save my thoughts to external paper storage.
I have over the decades tried every combination I can think up or have read about for designing things and I always land back in the same place, A4 5mm square pad (cheap but not the cheapest) and a really good (but not expensive) set of ballpoints in black/blue/red/green (and mostly it's black anyway).
There is something about the complete free form nature of pen and paper that makes me think better, I bought a nice ipad/pencil a few years ago thinking that would replicate it but it didn't, it was closer but not close.
If it's something I want to keep I just take a photo but vast majority of notes are transient anyway.
I rediscovered my university revision notes recently and I don't think it's a coincidence that the year I got my best grade was the year I went to town on very neat, carefully laid out lecture notes. Whether I was taking exquisite care over circuit diagrams or deciding which colour to make streamlines and velocity gradients, it helped bury the facts in my brain.
Exactly. I also use those colors.
I made a little system where blue diagrams are services I need to code and host like a nodejs server
green is stuff I need to host only like nginx, but no coding
red is external infra I can't control at all, like Anthropic's API.
And then I make drawings of how I connect them and in the end I just generate the code, so I go from pen and paper to claude code.
I spent an unreasonable amount of time recently thinking about how to color- (and symbol-) code my notes. I wanted to create a simple generally applicable system that I can use for any project.
For thinking, nothing beats handwriting. It's just the unconstrained freedom of movement in every direction, shape or form, and the lack of friction and distractions.
But then you sometime also want to organize or share stuff, so that's where digital comes in. I have an automation that takes a photo of my notes with my phone, uploads it to Obsidian, and OCRs it into markdown and tags it, so I basically end up with a "symbolic" 1-1 equivalent of analog handwriting + digital text
My system still evolves, but what I currently use is:
Color: - black = just "normal" text - red = critical/"must" - yellow = important/"should" - green = highlighted/"could"
Symbols: - bulletpoint [ ] task -> reminder
I also do brackets and underlines which translate to # heading1 and ## heading2 in markdown
I write by hand with a fountain pen with black ink + use 3 fineliners and small colored sticky flags.
Some observations: - yellow is not great on white paper due to bad contrast; it also rarely comes as ballpoint, but fineliner options are available - blue and green are visually very similar, to the point of kinda being redundant - for sharing with other people, the system must be as generally understandable as possible. Here the traffic norms such as ISO 7010 (https://en.wikipedia.org/wiki/ISO_7010) could be considered - also there's the aspect of accessibility, with a relevant portion of population being colorblind - The simplest most effective system seems to be: one main color (black or blue) and one highlight color (red I think works best), plus a couple of symbols such as bullet point, task, reminder that translate well into markdown
Cool! Im also thinking of turning them into markdown eventually with the idea to design software only on paper and then pass it off to claude after.
I really think we could go back to pen and paper fully for software design.
> A4 5mm square pad (cheap but not the cheapest) and a really good (but not expensive) set of ballpoints in black/blue/red/green
It sounds like you have some stationery recommendations!
I'm always in the market for really-good-but-not-expensive pens.
I use the "uni-ball - Jetstream SXN-210" (in 1mm), they are superb, don't smear, write smooth every time and are cheap enough if I lend one to someone and don't get it back I don't regret it, ergonomically they are comfortable as well, not a lot else to say about them beyond that - just a really well designed pen.
I could write an ode to the 5mm grid... Yellow Rhodia No. 16, I'm in love!
https://www.dickblick.com/products/rhodia-top-stapled-notepa...
Agreed, I used to role my eyes when my professors asked me to do things like pseudocode (which I still don't do too much) but the act of writing an outline for the general structure of a program always helped me get over the initial hump of starting a new project.
For me, writing is also a thinking tool, that is why I don't use AI to write. But using AI to generate ideas for writing? All the time :D (That way, I don't need to wait for that inspiration). I like the idea of scheduling it though!
One of the biggest thing we lose as we replace workflows with AI are the ancillary benefits of the process of writing (prose or code), design, or anything else.
You may produce an acceptable result faster with AI, but the deep understanding, questioning your own assumptions, and correcting your own errors only come from the actual writing process.
And these are not nice to haves because that cognitive context drives all downstream work.
"Writing is nature's way of letting you know how sloppy your thinking is."
--Richard Guindon
When i started writing i felt the pressure to always go from draft to a polished piece. i posted a couple things i worked semi hard on and received 0 views/reads because i don't really have much of an online presence at all.
From there i kinda just moved on and write purely for myself, and use it to up my own thoughts. Very rarely, something will receive a second pass days later, and thats just more to go back on my past thoughts and try to tighten them if i feel that is needed.
I tend to split my day into two parts, my work hours and my "unplug" hours, and the writing was mostly landing in those unplug hours as a way to sorta summarize my day. Last week or so I decided im moving that writing into the very first thing I do each day, and it has really been absurdly effective at getting new ideas flowing quickly. I used to depend on my later day brain that was somewhat "depleted", for lack of a better way to describe how i work, and from there that would kinda birth what i work on the next morning, with maybe some iteration while working.
Now instead I just do both, I open the day with ~500-1500 words, just anything that comes out, then i potentially get stuck, and ill maybe dump that to a model as just a modern rubber duck, not anything like "write for me, or finish this thought", just a kinda nudge to really toss back at me what im fully considering and then maybe i pivot back and chase down the most interesting of the topics for a few hundred more words, or get straight to work on what i thought was the most vital of all those ideas to chase while im fresh in the morning.
Writing is a good thinking tool.
But, to the claim of how to think in the age of LLMs, I think it’s worth contrasting the other approach, which is a dialog with the LLM; two things are needed IMO to keep things honest:
First, you need to keep asking the LLM to explain things, until you understand what it did. Lots of angles here; have the LLM make a design/product doc, write one-off architecture explainers, write throwaway data-viz tools to explain the data distribution. Then ask a fresh copy to adversarially review the core concepts. Try asking a different frontier model too.
Second, you need to push for “proof”, in other words, ask for demos and try it yourself. Imagine yourself a CEO or VC; how do you know if the output of an entire org is vaporware? One thing to beware is launching into a field where you don’t have expertise and can’t actually validate the output. This takes time of course, but for example if you write a stress test script and get CPU/QPS numbers, these complex validators are IME quite reliable when you apply the same rigor from 1) above to their construction.
I'd been a software engineer for 25 years before language models came into the scene. There were so many parts of the engineering world I had no idea about: Large cloud data processing, lower level device / GPUs, UI frameworks and design.
I kept my to my niche.
Now I can build anything, but as I'm building it, I try to understand at least how it works. So I'm constantly asking questions. In the process, I think of my prior engineering skill as somewhat childlike. As I thought I was a decent engineer, I now see that I was decent at some things, but far from the kind of breadth that the wide engineering world has.
I think the takeaway is that you should always be curious and build things. Whether it's writing or discussing with an language model, writing software by hand, or creating software with a model and understanding it, I think you have the same effect.
> Now I can build anything, but as I'm building it, I try to understand at least how it works. So I'm constantly asking questions.
This is the way!
AI is the best technology ever invented for avoiding learning things. It’s also the best technology ever invented for learning things.
I think it’s very important to be on your guard for accidentally glazing over your understanding, but if understanding is your goal, there are so many more tools for this now.
Before AI, even when working collaboratively and bumping into concepts I did not fully grasp, there was so much more friction to get answers. Sure I could Google and try my luck, or I could ask a colleague. But the LLMs have infinite patience, can build diagrams, and can answer weird questions I would be afraid to ask.
With writing I think it's the same thing.
Instead of asking the LLM to explain things, you can use it as a pen.
Let it organize and work with your thoughts, as you say, how you say, as you do it.
Once that is clear, it can then better create those other artifacts.
Words are a communicable but ultimately lossy map of reality, experience of which we cannot otherwise convey; if the goal, the final artefact is itself a map (e.g., a thesis), using writing at earlier stages is, in a way, beginning to sketch out said map.
I endlessly enjoyed the comparison with a Turning Machine. When I study someone's complex work, I do notes to keep several things in my sight scope: a piece of paper, PowerPoint slide or some fancy markdown text editor with syntax highlight. Seems like that this focusing is somewhat the natural way of building neuronal connections: if a logical piece is missing we have to use imagination to complete it and only then it nicely fits into our brain's "hologram" of the world.
There is a wonderful book "Exploring Writing" by John Langan. I found it in our library when I was in my PhD writing phase. I highly recommend it to anyone who has problems building text and who wants to learn how to master its structuring.
And for freewriting, sometimes when it gets really hard to move further: https://3am.freewriter.app/ If you set 5 min target duration, you will be awarded with a full version of the song!
I always carry a small Moleskine notebook, the pocket-sized hardcover with grid paper, for notes, thoughts, and whatever else comes to mind during the day. Each page is numbered, and I continue the numbering as I move from one notebook to the next. This allows quick reference to past ideas, thoughts, or notes I may have in an older notebook, or even a few pages back. It is the primary way I get thoughts to paper, and is the "entry point" to all my writing and thinking.
Notebooks do not crash, do not freeze, and need only you and a pen. My notebooks often remind me of this passage from "Zen and the Art of Motorcycle Maintenance":
> “For this you keep a lab notebook. Everything gets written down, formally, so that you know at all times where you are, where you’ve been, where you’re going and where you want to get. In scientific work and electronics technology this is necessary because otherwise the problems get so complex you get lost in them and confused and forget what you know and what you don’t know and have to give up.”
> Robert M. Pirsig, Zen and the Art of Motorcycle Maintenance
It's an attractive thought, this comparative look at FA and TM. However it is misleading to an extent.
A Turing Machine's capabilities exceed the FA's because it has a 'store of knowledge'. It is both recording and reading, repeatedly. There is erasure, too. So if we're aspiring to TM-hood, the guidance seems to be write, and re-read (your own thoughts), and re-evaluate, and re-write. This process is subject to the halting problem. /g
An observation as it relates to LLMs, anecdotal. I noticed that when I screenshot a handwritten scratchpad of my thinking, rather than just write a prompt with same thinking, they seem to be more “eager” to respond and more keen to understand the ideas. Maybe an illusion but if anyone else noticed this it would make two of us.
This has cause/effect backwards.
Writing for me anyway is more like a general anxiety release valve.
Thinking is we just do without external tools as we reassess frame to frame due to the world around us not being static; we develop object permanence of a tree for a while but are burning calories to replace on a much shorter timeline; whole lot of physical state still updating on shorter timelines than a tree; so our brains automatically always wipe and refresh the canvas as a natural consequence of relative entropy
We track these deltas and how to deal with them; writing helps deal with scenarios where there is no roof to fix to stay sheltered. What do I do with these thoughts with nowhere to go; I have food, I have shelter? Writing.
Reading Murat's blog around the start of the year introduced me to the idea of freewriting. I built a tool that I still use weekly for coming up with new ideas / exploring thoughts.
https://flow-six-henna.vercel.app https://github.com/sno6/flow
(Fully local btw)
I have seen the same thing happen in multiple domains. You may have identified a problem to solve, but the act of solving it, by writing it down, by writing a program, building a tool, making a device, or even carving from stone.
As you work on it the problem becomes more clear and seems to tell you what the actual shape of the problem is and you have to adapt your solution to that actual problem, even if it ignores certain aspects of the original idea.
Writing it down and trying to clarify it is one of the cheapest ways to get to the real answer.
A common refrain for teachers and writers is that they learned so much producing their work. You think you know it, and then you go to teach it and realize how much you didn't know.
I'm a big fan of marbled composition books. They're cheap and durable enough. You can write in them without a table. I have 5-6 of them going at once for various tasks. And somehow (for me) the ideas flow more freely longhand than they do at the keyboard. When full, you can cut the spine off with a box cutter and scan it. G4 compression loses the color, but makes really small PDFs.
It's easy to produce better written content than an LLM. And that's good for the reader and the writer.
I have been finding I spend more time writing when working with AI. While I do let it generate docs, it does not always do a great job and I end up reorganizing, reframing, or changing vocabulary. Those passes end up require rigorous thinking.
There has been some occasions where I cannot use AI as a thinking partner. It doesn’t have the imagination, intuition, or judgement. I have been relying more on my team in those cases.
This article reminds me of another method: grabbing paper and pencil and start doodling around to work out some ideas.
There's always this peculiarity regarding the fact that writing by hand, is not the same as writing by kbd. So I'm always confused which of these two actually authors, mean, though I agree - any sort of note taking is better than none.
For me I think the distinction is more about structure and rigidity than keyboard vs pencil. Sketching free-form notes and especially little diagrams is a really powerful thinking tool. I also like to sketch my thoughts in org mode, which has some advantages over paper (structural editing, ability to zoom in and out by folding subtrees) and some disadvantages (tree structure instead of arbitrary spatial structure, more friction for diagrams).
I have a really hard time using WYSIWYG editors for thinking-by-writing. I think it’s because it’s hard to see partially-formed structure so it feels less malleable.
Yes, me too! I like using sketch notes to think through things nonlinearly, then I make an outline in Org or use an AI to get the text from a sketch, and then maybe I'll refine the sketch so it can be an illustration that I embed back into Org. Structural editing is fantastic for reordering thoughts and converting back and forth between lists and paragraphs/sentences.
but then, from personal experience, once you master, and practice photoshop/illustrator-like software, it becomes not so hard to actually produce quite complex visual communication in little to no effort. it is questionable whether I'd ever be able to do so with pen and paper, though, of course, the feeling... the touch of it is second to none. Talking of practicality though, it makes me wonder. I think rather an important question is whether we also talk about writing/note taking as part of larger personal information management system. It is very likely having one (such own KB) matters even more than the medium itself.
You can get the best of both worlds by picking up a Wacom tablet or similar. It isn't exactly the same, but it does then mean you can doodle but with the power of the capable software.
definitely thinking about it every day, as the pain in my wrist, one resulting from touch and mouse use, is increasing with no end in sight. some pencil and paper based loop through the LLMs seems even more appealing.
Writing is a good (and necessary) tool for thinking.
Not all writing needs to be published.
Most shouldn't.
I have read many articles on the importance of writing, in management, in understanding things etc.
I always noticed that very few people actually write and very few people read.
Fast forward to now, LLMs happily write reams and even fewer people read.
Writing is not helpful just because it has Turing-tape memory or because it reduces noise in your signal to others.
It's mainly helpful in changing an exercise for youself into something useful for others, that requires you actually adopt their perspectives and migrate them to yours (or realize the error of your ways). You take on the capability not just to convey ideas but also to change the world.
Interestingly what has worked very well for me is that when I start a new project, I just have to start drawing a few boxes and connect them as I expect them to be communicating and with just an hour or so of doing this, I find myself ready to write code as if there was a block that got completely untangled with putting it on paper.
> Yet the only difference between a FA and a TM is that
> the TM, unlike the FA, has paper and pencil.
> Think about it.
The analogy doesn't works if you really think about it. A TM with a finite tape is equivalent to an FA. The fundamental difference between a TM and an FA is that a TM has infinite space.
(Yes, I know I missed the point and am not very fun at parties.)
No, this is a very good point - you need a supply of pen and paper, such that it feels effectively infinite. Buying a nice notebook or a nice pen is a trap because both feel finite; buying a box of decent pens and a carton of decent notepads takes you out of the scarcity mindset and unlocks the Turing machine like advantages mentioned.
I was thinking of the same. FSA can multiply within bounds. For FSA, the storage is the state itself and bounded. I agree with the sentiment but not a fan of analogies like this. We do have memory even if we don't write. Each organ adds to the memory imprint making it stronger. Visual, smell, taste, audio etc Each action adds more imprints. And this makes more sense to me than an analogy.
Another analogy which never made sense to me was improving by just 1% every day (1.01³⁶⁵).
Analogies do help explain phenomena I cannot even relate to at all, like telling me how life on Mars would be.
Ofcourse I'm not invited to parties as you can imagine, taking things literally.
You’re welcome to my party, both of you.
A big part of understanding something is getting used to it, and analogies help with getting over a mental hang up. They may also help with deciding which areas of the new space to explore first.
But we should be examining the limits of applicability, as we do for most cognitive tools, so we know when to switch from one tool to another. It also helps to combine such tools in a complementary manner.
Are you saying that 1.01^365 is too big? I think there are lots of things where if you are starting from near zero skill, that's not a crazy big number.
Not parent commentor, but My issue with 1.01^365 analogy is that it implies that somehow that skill multiplies itself. There is no obvious reason to believe the gain from _(skill acquisition, language learning, workout, whatever) is better modeled as 1.01^365 than 1+(0.1*365).
Being forced to take notes and write essays in school made me hate handwriting specifically and dislike writing in general. Now that I have a choice, I never use stationery or take notes for any purpose, and everything useful that I learned or did in my entire life has been done without writing being a part of the process.
Two things came from reading this. First, I feel inspired to write regularly again. Second, I need to read more of this blog. Thanks!
make it a habit of writing every day a couple sentences in the morning and another few in the evening, and if using a computer turn spellchecker off
Chain of thought, but for humans!
> Don't be discouraged by how bad the first draft was. McPhee says the way to do a piece of writing is three or four times over, never once. This is good news.
This also applies to code. Taking your agent's first draft as a sample and doing a reimplementation often yields an even cleaner version.
Once you do ship, the drafts' weaknesses provide codebase signal. Where should you shore up the architecture - or create more reusability, where is the documentation inadequate. Do you need to bake invariants deeper in so they're harder to violate (custom lints, stricter validation...)
And yeah. Writing is clutch too.
The human mind has a small L1 cache. High IQ people tends to have larger L1 cache, and don't understand when others can't follow their arguments. Not understanding that's it's dumb.
Writing as a thinking tool has become sort of a meme these days.
And, yes, I also generally believe it’s a good thing to use it as a tool. I try to do use it as often as I can.
But… I think it’s also important to note that there are - like with almost anything - pitfalls and downsides. I rarely see them discussed and thus want to bring them up:
For highly analytical people, writing may become counterproductive. Especially, when it is used to continuously elaborate thoughts rather than to challenge assumptions, test hypotheses or reach decisions.
It can lead to over-analyzing or pseudo-rationalizing.
Ask your favorite LLM for scientific references like this one: https://pubmed.ncbi.nlm.nih.gov/40137068/
Beyond all this, there's also tailoring your message to your audience.
In sales, you might call this finding "words that work" or "the pain points". AI will not know how to do this because it's getting a vibe check of the room.
You cannot merely be a sensory part for the AI unless you want to sound as cringe and out-of-touch as the majority of the training data (stale social media). Garbage in, garbage out.
I spent a lot of time thinking about it. AI converges on average which is very bland. The labs are trying to get out of it with RLFH which sadly just produces a different “edgy” average (Calude-speak).
But things get better if you instruct AI to speak in certain style. It’s still average but for the chosen style and thus less tone-deaf for a given use case.
So my thinking is - how much mileage can we get by introducing specific style guides? We can create a library of styles, and to get out of average for a specific guide individual companies and even people will maintain their own private patches from upstream.
Another option is to maintain history of your own writing and thus preserve your own voice. I feel like “averaging” of style is somehow un-ergonomic, in that in numbs the mind, and thus even diversity for the sake of diversity is valuable.
Words are most effective when they buck the trend in just the right way.
I'll use the conference room example for clarity, even though a lot of people still work remote these days.
Reading the room could be as subtle yet significant as reacting to the speaker that was on immediately before you. It could be reacting to the temperature of the room or the catered lunch everyone just had. It can be noticing or participating in or critiquing a peculiarity of the audience (hawaiian shirt fridays, etc.). It can also be a million other things and they'll never let you get away with choosing at random. You have to just "get it" or you're going to be ignored at best.
You're at once trying to both signal in-group status while also asserting your own identity and setting the tone for everything you'll be saying after. Remote work has its own idiosyncrasies, but regardless, you need a pulse for this kind of work. And, all this is just the introduction. What about emphasis choices in the rest of your writing/speaking?
Trying to pin that down is simply not logical and it's not just a matter of "adding" a handful of tricks unless you want to alienate everyone. Sometimes it's also the things you don't say or do that make the difference. You also have to accept that you can't please everyone. You could actually piss off everyone but the most important person in the room and still win. That might actually be the best way or the only way to win sometimes (as was the case for the American sales pitch of AI itself!).
The real world is messy and adding AI, or any particular process, to the mix just makes it worse. You cannot plan an accident, but you can avoid one.
That's how CoT got its name.
Wow is this the next billion dollar idea? Which startups are tackling this problem?
I drove my teachers nuts with my unconventional writing process. I start with a blank word processor document, and start making notes. Then I rearrange them, and finally I start turning it into an essay. My teachers stood over me, demanding that I stop misbehaving, and write an essay from start to finish. I could not do that. My brain does not work that way. I can not keep enough things in my head to organize a few pages that way.
---
I was once thrown out of a job interview after their secretary yelled at me for what seemed like 5 minutes because I was filling out a form from the shortest blank to the longest blank rather than from the top to the bottom. I had experienced such a trying day, that I just could not bring myself to fill out anything but the next shortest field. All the while, this crazy lady was yelling at me as I had to stand at her desk in front of a room of HR scum/manager scum and job applicants.
---
I classify linear thinking as one useful way to check if something makes sense, not the only way to think.
I sense a hint of malicious compliance. Did you also write all the vertical strokes, then the horizontal ones, then the loops? Vowels, then consonants? Write each letter in alphabetical order? Express your zipcode as a geometric series? You have a future in the go-slow bureaucracy of your choice! :-)
"Complexity in prose is usually a place to hide sloppy thinking, or a way to weasel out of taking a stance. If you can't say it plainly, you probably haven't thought it through yet."
This hit hard. Great article.
The place this surprised me most was before writing code. On a recent project I wrote a plain plan file first: numbered stages, what each one should do, what "done" means. Half the real decisions happened while writing it, not while building. Questions like "what happens when the external service is down" only showed up because I had to put the step into a sentence.
It also made the work resumable. Coming back after a few days, the file told me exactly where I was and why things were the way they were, which no amount of reading my own code did.
this is great but machines can think better than us now so we don't need to think
Im even saying this with some level of sincerity
Maybe better then you...