We can’t circumvent the work needed to train our minds
zettelkasten.de377 points by maksimur 3 days ago
377 points by maksimur 3 days ago
I was talking with somebody about their migration recently [0], and we got to speculating about AI and how it might have helped. There were basically 2 paths:
- Use the AI and ask for answers. It'll generate something! It'll also be pleasant, because it'll replace the thinking you were planning on doing.
- Use the AI to automate away the dumb stuff, like writing a bespoke test suite or new infra to run those tests. It'll almost certainly succeed, and be faster than you. And you'll move onto the next hard problem quickly.
It's funny, because these two things represent wildly different vibes. The first one, work is so much easier. AI is doing the job. In the second one, work is harder. You've compressed all your thinking work, back-to-back, and you're just doing hard thing after hard thing, because all the easy work happens in the background via LLM.
If you're in a position where there's any amount of competition (like at work, typically), it's hard to imagine where the people operating in the 2nd mode don't wildly outpace the people operating in the first, both in quality and volume of output.
But also, it's exhausting. Thinking always is, I guess.
[0] Rijnard, about https://sourcegraph.com/blog/how-not-to-break-a-search-engin...
I’ve tried the second path at work and it’s grueling.
“Almost certainly succeed” requires that you mostly plan out the implementation for it, and then monitor the LLM to ensure that it doesn’t get off track and do something awful. It’s hard to get much other work done in the meantime.
I feel like I’m unlocking, like, 10% or 20% productivity gains. Maybe.
10-20% productivity gains at the expense of making it grueling sounds like a recipe for burnout
The smart ones saw this early on.
The rest are just catching up to the reality now.
And that's how many people feel now.
I think it's a bad strategy
Burning out a substantial portion of the workforce for short term gains is going to cause way more long term decline than the short term gains are worth
I think the long term assumption is that the first path mentioned by trjordan mentioned above, where AI does all the work, is the goal. The second path is a necessary evil until the first path, which requires as yet unachieved improvements in AI (maybe approaching AGI, maybe not) becomes feasible. Burning out employees doesn't matter since they're still creating more value than they otherwise would, and they'll be replaced by AI anyway.
Agreed. Either that, or the task has really, really broad success parameters.
Yeah I think this is what I've tried to articulate to people that you've summed up well with "You've compressed all your thinking work, back-to-back, and you're just doing hard thing after hard thing" - Most of the bottleneck with any system design is the hard things, the unknown things, the unintended-consequences things. The AIs don't help you much with that.
There is a certain amount of regular work that I don't want to automate away, even though maybe I can. That regular work keeps me in the domain. It leads to epiphany's in regards to the hard problems. It adds time and something to do in between the hard problems.
> There is a certain amount of regular work that I don't want to automate away, even though maybe I can. That regular work keeps me in the domain. It leads to epiphany's in regards to the hard problems. It adds time and something to do in between the hard problems.
Exactly, some kinds of refactors are like this for me. Pretty mindless, kind of relaxing, almost algebraic. It's a pleasant way to wander around the code base just cleaning and improving things while you walk down a data or control flow. If you're following a thread then you don't even make decisions really, but you also get better acquainted with parts you don't know, and subconsciously get the practice holding some kind of gestalt in your head.
This kind of almost dream-like "grooming" seems important and useful, because it preps you for working with design problems later. Definitely formatting and style type trivia should absolutely be automated, and real architecture/design work requires active engagement. But there's a sweet spot in the middle.
Even before LLMs maybe you could automate some of these refactors with tools for manipulating ASTs or CSTs, if your language of choice had those tools. But automating everything that can be automated won't necessarily pay off if you're losing fluency that you might need later.
In my experience, a lot of the hard thinking gets done in my back-brain while I'm doing other things, and emerges when I take up the problem again. Doing the regular work gives my back-brain time to percolate; doing hard thing after hard thing doesn't.
Also at the end of the day, humans aren't machines. We are goopy meat and chemistry.
You cannot exclusively do hard things back to back to back every 8 hour day without fail. It will either burn you out, or you will make mistakes, or you will just be miserable.
Human brains do not want to think hard, because millions of years of evolution built brains to be cheap, and they STILL use like 10% of our daily energy.
I'd actually say that you end up needing to think more in the first example.
Because as soon as you realize that the output doesn't do exactly what you need, or has a bug, or needs to be extended (and has gotten beyond the complexity that AI can successfully update), you now need to read and deeply understand a bunch of code that you didn't write before you can move forward.
I think it can actually be fine to do this, just to see what gets generated as part of the brainstorming process, but you need to be willing to immediately delete all the code. If you find yourself reading through thousands of lines of AI-generated code, trying to understand what it's doing, it's likely that you're wasting a lot of time.
The final prompt/spec should be so clear and detailed that 100% of the generated code is as immediately comprehensible as if you'd written it yourself. If that's not the case, delete everything and return to planning mode.
> I'd actually say that you end up needing to think more in the first example.
Yes, but you are thinking about the wrong things, so the effort get spent poorly.
It is usually much more efficient to build your own mental model than to try to search for a solution that solves exactly what you need from externally. Without that mental model it is hard to evaluate whether the external solution even does what you want, so its something you need to do either way.
Depends how complex the task is. Sometimes I’m handed tasks so simple but tedious that AI has meant I can breeze through these instead of burning myself out on them. Sure, it doesn’t speed things up much in terms of time, but I’m way less burnt out at the end because it’s doing all the fiddly stuff that would tire me out. I suspect the tasks I get aren’t that typical though.
Yeah, I think if it's simple enough that you can understand all the code that's generated at a glance, then it's fine. There are definitely tasks that fit this description—my comment was mainly speaking to more complex tasks.
regarding #2: "Automate the dumb/boring stuff", I always think of the big short when Michael Burry said "yes I read all the boring spreadsheets, and I now have a contrary position". And ended up being RIGHT.
For example, I believe writing unit tests is way too important to be fully relegated to the most junior devs, or even LLM generation! In other fields, "test engineer" is an incredibly prestigious position to have, for example "lead test engineer, Space X/ Nasa/etc" -- that ain't a slouch job, you are literally responsible for some of the most important validation and engineering work done at the company.
So I do question the notion that we can offload the "simple" stuff and just move on with life. It hasn't really fully worked well in all fields, for example have we really outsourced the boring stuff like manufacturing and made things way better? The best companies making the best things do typically vertically integrate.
I stay at the architecture, code organization and algorithm level with AI. I plan things at that level then have the agent do full implementation. I have tests (which have been audited both manually and by agents) and I have multiple agents audit the implementation code. The pipeline is 100% automated and produces very good results, and you can still get some engineering vibes from the fact that you're orchestrating a stochastic workflow dag!
The problem with LLMs is that they are not good enough to do the dumb stuff by themselves. and they are still so dumb that they will bias you once you have to intervene.
But this is the idea behind compilers, type checkers, automated testing, version control, and etc. It's perfectly valid.
> You have to remember EVERYTHING. Only then you can perform the cognitive tasks necessary to perform meaningful knowledge work.
You don't have to remember everything. You have to remember enough entry points and the shape of what follows, trained through experience and going through the process of thinking and writing, to reason your way through meaningful knowledge work.
"It is requisite that a man should arrange the things he wishes to remember in a certain order, so that from one he may come to another: for order is a kind of chain for memory" – Thomas Aquinas, Summa Theologiae. Not ironically I found the passage in my Zettelkasten.
It's weird to read this from zettelkasten.de, given that the method is precisely about cultivating such a graph of knowledge. "Knowing enough to begin" seems to me to be the express purpose of writing and maintaining a zettelkasten and other such tools.
I would say they mean being able to recall, not having everything at once. It’s being able to answer the 5 why’s.
I arrange my code to follow a certain order, so that I can get my head back into a given module quickly. I don't remember everything; there's too much over the weeks, months, and years. But I can remember enough to find what I need to know if I structure it properly. Not unlike, you know, a Zettlekasten.
Isn't a Zettlekasten, structured so you can easily go back to a note easily?
This is task-specific. Consider having a conversation in a foreign language. You don't have time to use a dictionary, so you must have learned words to be able to use them. Similarly for other live performances like playing music.
When you're writing, you can often take your time. Too little knowledge, though, and it will require a lot of homework.
There might be words I don’t use or chords I don’t know. It doesn’t matter though because part of expertise is being able to consult a reference and go “of course”, implement it, and keep moving.
Actually this is how LLMs (with reasoning) work as well. There is the pre-training which is analogous to the human brain getting trained by as much information as possible. There is a "yet unknown" threshold of what is enough pre-training and then the models can start reasoning and use tools and the feedback from it to do something that resembles to human thinking and reasoning. So if we don't pre-train our brains with enough information, we will have a weak base model. Again this is of course more of an analogy as we yet don't know how our brains really work but more and more it is looking remarkably aligned with this hypothesis.
I always tell people that I don't remember all the answers, only where to find them.
Just to be clear, are you saying that to know something:
1- You may remember only the initial state and the brain does the rest, like with mnemonics
2- You may remember only the initial steps towards a solution, like knowing the assumptions and one or two insights to a mathematical proof?
I'd say a Zettlekasten user would agree with you if you mean 1
Of course you have to remember everything. Your brain stores everything, and you then get to add things by forgetting, but that does not mean you erase things. The brain is oscillatory, it works somehow by using ripples that encode everything within differences, just in case you have to remember that obscure action-syntax...a knot, a grip, a pivot that might let you escape death. Get to know the brain, folks.
Interesting take. I respectfully differ. IIRC, Feynman said something akin to my POV:
Brains are for thinking. Documents / PKM systems / tools are for remembering.
IOW: take notes, write things down.
FWIW I have a degree in cognitive psychology (psychobiology, neuroanatomy, human perception) and am an amateur neuroscientist. Somewhat familiar w/ the brain. :)
Feynman wasn't a neurobiologist.
I'd read Spontaneous Brain by Northoff (Copernican, irreducible neuroscience) or oscillatory neurobiology Buzsaki.
The brain is lossless.
I would agree that external forms of memory are evolutionarily progressive, that ability to utilize the external forms requires a lossless relationship.
Once we grasp the infinitely inferior external of arbitrariness (symbols words) are correlated through superior, lossless, concatenated internals (action-neural-spatial-syntax), until we can externalize that direct perception, the externals are deeply inferior, lossy forms.
But taking notes and writing ideas out requires that we think them through...which we usually don't do otherwise. This has been a commonplace of the intellectual life for centuries.
Words and thoughts are wholly separate. Notes aren't the direct results of perception, they are more like sportscasters reading the mind of pitchers. Notes point to thoughts or observations, they aren't the thoughts themselves.
“We refute (based on empirical evidence) claims that humans use linguistic representations to think.” Ev Fedorenko Language Lab MIT 2024
I did not say that my brain uses linguistic representations internally when I think; I said that the process of turning my ideas into words helps me think.
Actually you said "writing ideas out requires that we think them through" and this isn't what's happening in brains. In actuality, words interfere with our ability to think.
"The brain is lossless."
Nothing is lossless.
Fourier transforms are lossless. If it entered the oscillations of senses, it's still there in your brain. You may never need it, but every action is detailed by difference.
A Fourier transform is just a change of coordinates. It has nothing to do with the signal per se. If you have a signal which was measured or recorded with finite precision (as any signal must be) then the fourier transform (as a pure mathematical object) simply preserves the same amount of loss that the original signal had. But, in fact, to do that, we would need to do the transform on hardware that could represent real numbers. This hardware does not exist in computers or in your brain, and so a fourier transform is lossy in that case. Still, the idea that your brain encodes all information in oscillations is not accurate - your temporary electrical activity can be substantially disrupted without you losing your memories, suggesting very strongly (to put it mildly) that some of your memories are encoded chemically and physically in changes to the connectivity between neurons that do not depend on persistent electrical activity. That encoding scheme must be lossy.
Fourier transforms are simply ways of interpreting. The senses restrict, the existing memory/engram/affinity of those senses: we don't yet know how much is lost, whether any of it is lost. In the schema, we must assume their inputs are preserved in some way, (study those with photographic memory, or the eidetic as example) and it is assumed preservation is chemical in nature that represents or co-represents the oscillations, yet we're not sure what exactly this is as of 2024 (I haven't finished reading this year's). Of course there are disruptions, and sleep - you do not remember the sounds around us when asleep. But if you recorded an event, it is still there. By lossless, I am referring to the idea there is no true separation between the world and the brain. While the senses mitigate, the fields interact. Would the toroidal manifold of the entorhinal cortex exist without a Copernican-Mach universe? Of course not. In essence, as Northoff illustrates on his last book's cover, the universe is within us. He is neither being poetic nor ironic. Rather paradoxical. Brain's are oscillatory Copernican bulbs, the design elements of the brain reflect the Machian idea that everything affects.
Without direct perception, and using such poor tools as symbols and narratives to externalize memory, we're deeply impoverished as to the nature of memory and our ability to access it. But once we have a better grasp of the neuronal units, spatial-syntax, we will unlock every memory.
https://pmc.ncbi.nlm.nih.gov/articles/PMC10500127/
Also to consider are the shapes and phases between oscillation. "It’s high dimensional complexity; the mind is an attractor in high dimensional phase space formed between neural oscillators." Emergent properties are not reducible to their constituent parts.
fourier transforms are lossless, but what impleemntation are you refering to that losslessly implements a fourier transform?
to my knowledge practical fourier transforms set a number of sine waves they will calculate for, and a window of time to look at. these limitations result in loss.
but, just taking the brain, at some point the person will die and decompose. how are you gonna get the oscillations back out of the rotted flesh? there has to be some form of loss to the brain
We only need brains when we're alive, so extracting the points isn't required.
In terms of brains, the math is used to model the irreducible occurrences in brains - that everything is still in there. So the math only gives us a window into the complexity. Brains don't compute or calculate necessarily. As an analog, or analoga of differences, it never has to exclude, or experience loss.
For the details: Rhythms of the Brain or Unlocking the Brain both volumes.
Popsci books tend to be horseshit.
Reading one does not make YOU a neurobiologist.
They're not popsci books. I'm a co-lead developer on a project with neurobio consultants, so I better know wtf they're talking about.
I don't think that the point of the article was "you are dumb if you don't remember absolutely everything".
The point, I believe, was that the more you remember, the better you can think. As in you should strive to remember stuff, and not just be lazy and rely on LLMs. I agree with that.
A bit like the memory palace. One memory leads to another. Not random-access.
You only need the initial seed to restore the full state, provided you can reason your way from there. If you haven't applied yourself to problem solving, then perhaps you might need to memorize the full state.
Executing on meaningful knowledge work also might require many different paths, depending on the context and the environment. To me it's more about the method of inquiry and how you begin than it is the specific content. Sure, more individual facts help to guide that inquiry, but at any given moment you're only truly going to be able to recall a subset of those.
> You have to remember EVERYTHING. Only then you can perform the cognitive tasks necessary to perform meaningful knowledge work.
If humans did not have any facilities for abstraction, sure. But then "knowledge work" would be impossible.
You need to remember some set of concrete facts for knowledge work, sure, but it's just one—necessary but small—component. More important than specific factual knowledge, you need two things: strong conceptual models for whatever you're doing and tacit knowledge.
You need to know some facts to build up strong conceptual models but you don't need to remember them all at once and, once you've built up that strong conceptual understanding, you'll need specifics even less.
Tacit knowledge—which, in knowledge work, manifests as intuition and taste—can only be built up through experience and feedback. Again, you need some specific knowledge to get started but, once you have some real experience, factual knowledge stops being a bottleneck.
Once you've built up a strong foundation, the way you learn and retain facts changes too. Memorization might be a powerful tool to get you started but, once you've made some real progress, it becomes unnecessary if not counterproductive. You can pick bits of info up as you go along and slot them into your existing mental frameworks.
My theory is that the folks who hate memorization are the ones who were able to force their way through the beginner stages of whatever they were doing without dull rote memorization, and then, once there, really do not need it any more. Which would at least partly explain why there are such vehement disagreements about whether memorization is crucial or not.
I have seen multiple posts about this "remembering EVERYTHING", and I think they miss the point. Also they don't quote the context:
> So, coming back to the initial starting point that “you don’t have to remember anything”. The opposite is true. You have to remember EVERYTHING.
I see it like this: it is absolutely wrong to think that you don't have to remember anything. In fact, ideally you would remember everything. The more you remember, the better you can think. Now in practice, it's impossible to remember absolutely everything, so we should strive to remember as much as we can. And of course we need to be clever in how we select what we remember (but that seems obvious).
The point is really that it is common to say "it's useless to remember it because you can ask your calculator or an LLM", and the article strongly disagrees with that.
> More important than specific factual knowledge, you need two things: strong conceptual models for whatever you're doing and tacit knowledge
And the more experience with computers I get, the more I realize that there's actually not that many pure unique and mutually orthogonal _concepts_ in computer science and software engineering. Yes, a competent engineer must know, feel, live these concepts, and it takes a lot of work and exposure to crystallize them in the brain from all the libraries, books, programs, architectures one has seen. But there's not a lot of them! And once you are intimate with all of them, you can grok anything computer-related quickly and efficiently: because your brain will just wuickly find the "coordinates" of that thing in the concept space, ans that's all you'll have to understand and recall later.
I've been having conversations about this topic with friends recently and I keep coming back to this idea that most engineering work, which I will define as work that begins with a question and without a clear solution, requires a lot of foundational understanding of the previous layer of abstraction. If you imagine knowledge as a pyramid, you can work at the top of the pyramid as long as you understand the foundation that makes up your level, however to jump a level above or below that would require building that foundation yet again. Computer science fits well into this model where you have people at many layers of abstractions who all work very well within their layer but might not understand as much about the other layers. But regardless of where you are in the pyramid, understanding ALL the layers underneath will lead to better intuition about the problems of your layer. To farm out the understanding for these things will obviously end up having negative impact not just on overall critical thinking, but on the way we intuit how the world works.
> If you can’t produce a comprehensive answer with confidence and on the whim the second you read the question, you don’t have the sufficient background knowledge.
While the article makes some reasonable points, this is too far gone. You don't need to know how to "weigh each minute spend on flexibility against the minutes spent on aerobic capacity and strength" to put together a reasonable workout plan. Sure, your workouts might not be as minmaxed as they possibly could be, but that really doesn't matter. So long as the plan is not downright bad, the main thing is that you keep at it regularly. The same idea extends to nearly every other domain, you don't need to be a deep expert to get reasonably good results.
The US is, however, learning exactly what happens when rationality is not part of the equation. This is all a dance around what is a "fact" and how to string facts into a reasoning model that lets you predict or confirm other potential facts, etc...
It's simply different people we're talking about. Certain personalities are always going to gravitate to the "search for reason" model in life rather than "reason about facts".
At least in the field of sports science and exercise physiology, we have very little in the way of facts. Much of what we once thought were facts have been disproven or at least called into question by later research. So we need to be humble, and very circumspect in what we label as a "fact".
Yeah, so much of this is not about facts as much as judgment: knowing what are the most themes and factors to making a system work well. What are the parts that, if you get right, everything else will follow. Knowing what areas it's okay, even beneficial, to rediscover instead of trying to plan ahead of time.
Do you have any go-to examples of “facts” that were disproven in this field?
Zone 2 training is a good use of time for amateur endurance athletes.
https://www.youtube.com/watch?v=QQnc6-z7AO8
Post exercise muscle soreness is caused by lactic acid buildup.
https://my.clevelandclinic.org/health/body/24521-lactic-acid
Icing is an effective treatment for minor musculoskeletal injuries.
https://www.yalemedicine.org/news/rice-protocol-for-injuries
Weight training for children will stunt their growth.
https://peterattiamd.com/belindabeck/
Stretching before exercise prevents injuries.
https://mcpress.mayoclinic.org/nutrition-fitness/does-stretc...
I could go on and on but I think you get the point. Almost everything you hear in this field is opinions, extrapolations, and educated guesses rather than anything we could really call a "fact".
Like what? There are some basics that have been studied and represent useful approximations. The general statement that your body "makes specific adaptations to imposed demands" seems to hold no matter what you do. There seems to always be some debate about what demands to impose to get a specific adaptation. For example, people have a very diverse array of opinions about how and when to stretch to achieve certain kinds of flexibility and if you do some reading you will find that these opinions follow from a body of work in a specific activity and that they don't always translate well to other activities.
I am sympathetic to memory-focused tools like Anki and Zettelkasten (haven't used the latter myself, though) but I think this post is a bit oversimplified.
I think there are at least two models of work that require knowledge:
1. Work when you need to be able to refer to everything instantly. I don't know if this is actually necessary for most scenarios other than live debates, or some form of hyper-productivity in which you need to have extremely high-quality results near-instantaneously.
(HN comments are, amusingly, also an example – comments that are in-depth but come days later aren't relevant. So if you want to make a comment that references a wide variety of knowledge, you'll probably need to already know it, in toto.)
2. Work when you need to "know a small piece of what you don't remember as a whole", or in other terms, know the map, but not necessarily the entire territory. This is essentially most knowledge work: research, writing, and other tasks that require you to create output, but that output doesn't need to be right now, like in a debate.
For example, you can know that X person say something important about Y topic, but not need to know precisely what it was – just look it up later. However, you do still need to know what you're looking for, which is a kind of reference knowledge.
--
What is actually new lately, in my experience, is that AI tools are a huge help for situations where you don't have either Type 1 or Type 2 knowledge of something, and only have a kind of vague sense of the thing you're looking for.
Google and traditional search engines are functionally useless for this, but asking ChatGPT a question like, "I am looking for people that said something like XYZ." This previously required someone to have asked the exact same question on Reddit/a forum, but now you can get a pretty good answer from AI.
The AI can also give you pretty good examples of "kind" that you can then evaluate. I've had it find companies that "do X" and then used those companies to understand enough about what I am or am not looking for to research it myself using a search engine. The last time I did this I didn't end up surfacing any of what the AI provided. It's more like talking to the guy in the next cubicle, hearing some suggestions from them, and using those suggestions to form my own opinion about what's important and digging in on that. You do still have to do the work of forming an opinion. The ML model is just much better at recognizing relationships between different words and between features of a category of statements, and in my case they were statements that companies in a particular field tended to make on their websites.
> What is actually new lately, in my experience, is that AI tools are a huge help for situations where you don't have either Type 1 or Type 2 knowledge of something
IMO, this is the whole point of the article: AI tools "help" a lot when we are completely uninformed. But in doing that, they prevent us from getting informed in the first place. Which is counter-productive in the long term.
I like to say that learning goes in iterations:
* First you accept new material (the teacher shows some mathematical concept and proves that it works). It convinces you that it makes sense, but you don't know enough to actually be sure that the proof was 100% correct.
* Then you try to apply it, with whatever you could memorise from the previous step. It looked easy when the teacher did it, but when you do it yourself it raises new questions. But while doing this, you memorise it. Being able to say "I can do this exercise, but in this other one there is this difference and I'm stuck" means that you have memorised something.
* Now that you have memorised more, you can go back to the material, and try to convince yourself that you now see how to solve that exercise you were stuck with.
* etc.
It's a loop of something like "accept, understand, memorise, use". If, instead, you prompt until the AI gives you the right answer, you're not learning much.
"IMO, this is the whole point of the article: AI tools "help" a lot when we are completely uninformed. But in doing that, they prevent us from getting informed in the first place. Which is counter-productive in the long term."
Great way of framing it - simple and cuts straight to the heart of the issue.
Pilots have both checklists that they can follow without memorizing, but also memory items that have to be performed almost instinctively if they encounter the precondition events.
Live performance (like conversation or playing music) often relies on memory to do it well.
That might be a good criteria for how much to memorize: do you want to be able to do it live?
It’s the same with math. A lot of people say they don’t need to be able to do basic arithmetic because they can use a calculator. But I think that you can process the world much better and faster if at a minimum you have some intuition about numbers and arithmetic.
It’s the same with a lot of other things. AI and search engines help a lot but you are at an advantage if at least you have some ability to gauge what should be possible and how to do it.
I used to find it weird how many people would make an excel formula on data they couldn't intuitively check. Like even basic level 'what percentage increase is a8 from a7' - they enter a formula then don't know if it's correct. I always wrote formulas on numbers I can reason with. If a8 is 120, and a7 is 100 you can immediately tell if you've gone wrong. Then you change for 1,387 and 1,252 and know it's going to be accurate.
People do the same with AI, ask it about something they know little about then assume it is correct, rather than checking their ideas with known values or concepts they might be able to error check.
With or without calculator some people have an aversion to calculation and that's the problem in my opinion. How much bullshit you can refute with back of the envelope calculations is remarkable.
This, and knowing by heart all the simple formulas/rules for area/volume/density and energy measurements.
The classic example being pizza diameter.
The actual central point is that the brain requires conditioning via experience. That shouldn't be controversial, and I can't decide if the general replies here are an extended and ironic elaboration of his point or not.
If you never memorize anything, but are highly adept at searching for that information, your brain has only learned how to search for things. Any work it needs to do in the absence of searching will be compromised due to the lack of conditioning/experience. Maybe that works for you, or maybe that works in the world that's being built currently, but it doesn't change the basic premise at all.
I agree with the point being made, even if it is taken to an extreme. I would say you don't need to remember everything, but you do need to have been exposed to it. Not knowing what you don't know is a huge handicap in knowledge work.
“Try to learn something about everything and everything about something.”
Before the internet we asked people around us in our sphere. If we wanted to know the answer to a question, we asked, they made up an answer, and we believed it and moved on.
Then the internet came, and we asked the internet. The internet wasn't correct, but it was a far higher % correct than asking a random person who was near you.
Now AI comes. It isn't correct, but it's far higher % correct than asking a random person near you, and often asking the internet which is a random blog page which is another random person who may or may not have done any research to come up with an answer.
The idea that any of this needs to be 100% correct is weird to me. I lived a long period in my life where everyone accepted what a random person near them said, and we all believed it.
If you are asking random people, then your approach is incorrect. You should be asking the domain experts. Not gonna ask my wife about video games. Not gonna ask my dad about computer programming.
There, I've shaved a ton of the spread off of your argument. Possibly enough to moot the value of the AI, depending on the domain.
This all assumes you have experts that you can talk to. But they might be difficult to find or expensive to hire. You wouldn't want to waste your lawyer's time on trivia.
That is why experts often publish books and articles, which is then corrected by other experts (or random people if it’s a typo). I’ve read a lot of books and I haven’t met any of their authors. But I’ve still learned stuff.
Yep. At that point you're doing research, and become familiar enough with the literature to know what's right is work.
Much like with Wikipedia, using AI to start on this journey (rather than blindly using quick answers) makes a lot of sense.
Most of us don't have the domain expert in range, particularly pre-internet. I was merely suggesting you ask the most expert of the domain you have access to and work your way up the tree of knowledge from there.
However, the sibling commenter about books, journals, etc., is also an excellent suggestion.
Before the internet, I didn't have the phone number of domain experts to just call and ask these questions. perhaps you did. For a lot of us, it was an entirely foreign experience to have domain experts at your finger tips.
How is an LLM making stochastic inferences based on aggregations of random blog pages more likely to be correct than looking things up on decidedly non-random blog pages written by people with relevant domain knowledge?
Is the above comment a genuine question? I’m concerned it’s a rhetorical question that isn’t really getting to the heart of the matter; namely, what is the empirical performance? One’s ability to explain said performance doesn’t always keep up.
How about we pick an LLM evaluation and get specific? They have strengths and weaknesses. Some do outperform humans in certain areas.
Often I see people latching on to some reason that “proves” to them “LLMs cannot do X”. Stop and think about how powerful such a claim has to be. Such claims are masquerading as impossibility proofs.
Cognitive dissonance is a powerful force. Hold your claims lightly.
There are often misunderstandings here on HN about the kinds of things transformer based models can learn. Many people use the phrase “stochastic parrots” derisively; most of the time I think these folks are getting it badly wrong. A careful reading of the original paper is essential, not to mention follow up work.
I'm not making a blanket statement against LLMs for all use cases. I'm certain that LLMs are, for example, much more performant at indexing already-curated documents and locating information within them than humans operating manually are.
What I'm skeptical about isn't LLMs as a utilitarian tool to enhance productivity in specific use cases, but rather treating LLMs as sources of information in their own right, especially given their defining characteristic of generating novel text through stochastic inference.
I'm 100% behind RAG powering the search engines of the future. Using LLMs to find reliable sources within the vast ocean of dubious information on the modern internet? Perfect -- ChatGPT, find me those detailed blog posts by people competent in the problem domain. Asking LLMs to come up with their own answers to questions? No thanks. That's just an even worse version of "ask a random person to make up an answer on the spot".
Descartes' brief rules for the direction of the mind [1] is pertinent here, as it articulates beautifully what it means to do "thinking" and how that relates to "memory".
Concepts have to be "internalized" into intuition for much of our thinking, and if they are externalized, we become a meme-copy machine as opposed to a thinking machine.
[1] https://en.wikipedia.org/wiki/Rules_for_the_Direction_of_the...
The author makes a lot of bold claims and I don’t take his main one serious re: remembering everything. I think he’s being intentionally hyperbolic. But the gist is sound to me, if you can put one together. He needs an editor.
> To find what you need online, you require a solid general education and, above all, prior knowledge in the area related to your search. > > [...] > > If you can’t produce a comprehensive answer with confidence and on the whim [...] you don’t have the sufficient background knowledge. > > [...] > > This drives us to one of the most important conclusions of the entire field of note-taking, knowledge work, critical thinking and alike: You, not AI, not your PKM or whatever need to build the knowledge because only then it is in your brain and you can go the next step. > > [...] > > The advertised benefits of all these tools come with a specific hidden cost: Your ability to think. [This passage actually appears ahead of the previous one–ed.]
This is best read alongside: https://news.ycombinator.com/item?id=45154088
One thing that I like is that things are much easier in person. When someone shows me an AI overview they just googled on their phone, I can say "I don't think that's true." Then we can discuss. The more we talk about the subject, the more we develop our knowledge. It's not black and white.
But online? @grok is this true?
Before the internet I siloed knowledge that I could lookup to books. Don't worry, the kids will be ok.
Indeed, I thought that "decades old" sounds like an underestimate there: Socrates is said to have criticized writing for letting people to not train their memory, so that would be millennia by now. Though of course it is possible that the article's author would not agree with that, and would have a beef with more easily searchable content only, like the people who criticized tables of contents. I do not mean that they were all wrong though: probably the degree to which knowledge is outsourced matters, maybe some transitions were more worthwhile than others, and possibly something was indeed lost with those.
>Socrates is said to have criticized writing for letting people to not train their memory
And he likely had a point. What I recently noticed is that my father, who had very little formal education but happened to have very old-school teachers who hammered him with memorizing a lot of poetry, which he continued in adult life, is more verbally fluent somehow than a lot of young kids who don't have command of grammar any more. ("would of")
Granted they likely don't write or read much either but directionally if you keep outsourcing mental work, you degrade. When I studied Japanese I liked the term a teacher had for his defense of memorizing Kanji by handwriting, which he called "neuro-muscular". Like playing scales on the guitar or piano there is something that keeps you snappy in memorization and rote practice that goes away if you only passively search.
Sorry, kids lack the foundational ability to remember, reason, imagine because their phones cauterize their basic intelligence foundations in sharp wave ripples: navigation, adventurous short-cuts, vicarious trial and error, these are the basis for memory consolidation. And we build this developmentally until we are 16 or so. Once we offload this dev to phones, we are essentially unintelligent buffoon, lacking the basis for knowledge. The kids are DOA.
No, you're just saying that. There's no evidence that using phones makes teenagers stupid.
Anyone who understands the development of intelligence and creativity is directly linked to the allocortex's ability to navigate freely, use vicarious trial and error, to invent novel short-cuts, built from both egocentric (landmark memories) and allocentric (extra body mapping) up until around the age of 16 in order to develop the basics of memory consolidation can take anecdotal evidence of kids that can't take a walk without a cellphone's help and extrapolate that these kids lack critical thinking skills.
It's elementary deduction from basic learning practices we've known since O'Keefe in 1973.
Given that the current generation of kids missed out on several crucial years of socializing due to Covid and were forced to find community online, I'm skeptical of arguments that point to poorly-socialized kids and say, "it must be the phones." Even if this was based on real data and not a hodgepodge of anecdotes, the phones themselves would not be my #1 suspect.
> allocortex's ability to navigate freely
If your allocortex is navigating freely, something has gone badly wrong. Put it back under the neocortex where it belongs and seek immediate help from a neurologist.
Sorry, that's a narrative argument, lack of covid socializing, the sciences tell us otherwise. Phones correlate to cog decline.
> Phones correlate to cog decline.
Do they? Let's check your source.
> To investigate this hypothesis, participants aged 20–34 perform a concentration and attention test in the presence and absence of a smartphone. The results of the conducted experiment imply that the mere presence of a smartphone results in lower cognitive performance, which supports the hypothesis of the smartphone presence using limited cognitive resources.
So, no. The presence of your smartphone on the desk in front of you is distracting, but that distraction goes away if you remove the smartphone. That's not "cognitive decline."
> Sorry, that's a narrative argument
No, that's me pointing out a competing plausible hypothesis. I'm not saying Covid is necessarily responsible for your anecdotal incidents; I'm saying that until you can prove Covid wasn't responsible, you have no standing to state conclusively that phones were.
We're in an attentional crises that isn't because kids didn't go to school for two years.
It's destroying their ability to experience reality as paths, free navigation, vicarious trial and error, all of this is fundamental to memory consolidation: the brain's fundamental unit: action-syntax in memory, is built from non-screen topological integrations of landmark and allocentric experiences. Phones destroy this.
https://pmc.ncbi.nlm.nih.gov/articles/PMC6059409/
https://www.nature.com/articles/s41598-022-20922-0
https://onlinelibrary.wiley.com/doi/10.1002/brb3.70656
https://www.mdpi.com/2254-9625/15/6/98
https://pubmed.ncbi.nlm.nih.gov/40255102/
https://link.springer.com/article/10.1007/s00246-025-03862-0
https://www.mdpi.com/2227-9067/12/4/503
https://pubmed.ncbi.nlm.nih.gov/40172268/
https://pubmed.ncbi.nlm.nih.gov/40173157/
https://www.nature.com/articles/s41390-025-04024-x
https://www.sciencedirect.com/science/article/abs/pii/S00223...
“A growing body of evidence has found that children’s brains can structurally and functionally change due to prolonged media multitasking, such as diminished gray matter in the prefrontal cortex, where attentional control and complex decision making abilities reside, among other really important skills, like the development of empathy and understanding nonverbal social communication,”
Your first study says smartphone use is associated with conduct problems and hyperactivity in 6-year-olds; no teens or stupidity in sight.
Other studies discuss distraction, cyberbullying, bad diet/poor exercise, toddlers' sensory processing abilities, anxiety, and lost sleep. None of this covers stupidity, and much of it is not about teens.
The most interesting study you cited finds structural brain differences in preschool-aged children who spent more time on screens. It's still a stretch to make claims about the intelligence of teenagers based on the fact that babies who spend too much time on screens are, at age 3, less developed.
There's also a magnitude problem. Even if we assume smartphones do have some cognitive effect on teens, how can we know it's the only or largest one? You can't attribute anecdotes about kids being dumb to the presence of smartphones.