I wouldn't call this nitpicking. This is how people who are careful with money think. I learned embarrassingly late to stop justifying purchases by making predictions about future returns. I treat everything as having zero value as soon as I purchase it. Thinking otherwise is, for me, always a dangerous rationalization -- always a craving that's trying to outmaneuver sense.
> The reality is that a human will learn, given any materials including LLMs, but only if they truly desire to learn. We've had MOOCs, gigantic libraries, all full of free information. You can obtain a PhD level understanding in any technical field of your choice today just by consistently going to the library and consistently applying yourself.
Not true. In every field there is guild knowledge that a person can't acquire from a library. In technical disciplines PhD-level knowledge requires experience in collaboration, research, and frequently lab work, which is impossible to acquire without access to a lab -- or just direct experience with research methods, whatever those may be. Reading papers and absorbing information aren't enough. PhD-level knowledge comes from the process of writing and doing original work.
> The reality is that a human will learn, given any materials including LLMs, but only if they truly desire to learn.
Also not true. We require kids to go to school partly because exposure to the environment and work inculcates skills regardless of whether kids want to do the work -- and regardless of whether they want to learn.
LLMs are damaging to students partly because they provide an escape hatch from that work and thereby prevent kids from acquiring skills.
Think of it this way: most people who want to be healthy and eat a healthy diet still find easy junk food tempting. What they want does not change the temptation, because the body and brain gravitate towards easy, cheap fulfillment of basic drives.
People facing challenging tasks, similarly, are tempted to take measures that reduce the amount of effort they require. The availability of tools that reduce the required effort also help shape a person's understanding of the value of the challenge and the work: "why should I do this hard task when I have a tool that can do it for me?" You and I know the answer to this question when we're discussing something like writing an essay or solving a problem in a math or programming class. Students frequently don't. They are by definition ignorant. Children, moreover, lack maturity. Their brains are less capable of resisting the easy path than an adult's. That's partly why parenting is important: parents provide boundaries and limits that kids need but won't and can't provide for themselves.
Sometimes people, especially kids, really do need to be dragged, kicking and screaming, through something in order to receive the benefits it offers. Being dragged through it sometimes convinces a person of its value and benefits. In a kid's case, there's a decent chance that the experience will improve executive function, shape expectations in a healthy way, inculcate grit, and become appreciation -- or at least habit.
I would not have written essays on my own as a student in secondary school. My English teachers had to provide that structure for me and impose the demand. But LLMs make it much more difficult to impose the demands, and kids are ill protected against the temptations of the cognitive equivalent of junk food, but an order of magnitude worse and more damaging.
The comment isn't a diatribe, and your experience as a teacher and advisor does not make your argument correct, let alone make your request any less a nonsequitur. If I disagree with you, I will always, I promise, say so. Asking that I 'please refrain' is just supercilious and silly.
It's also arguably a violation of Hacker News' guidelines:
> The LLM has terrible judgment but an encyclopedic knowledge
Bingo. Claude can bang out nearly correct code when I give it an idea, but it doesn't have the idea and repeatedly misjudges both how much work remains and what kind.
On the other hand, I don't know all the ins or outs of macro expansion in yaml at compile time or when and where macros run, enabling us to conaume their results elsewhere in the yaml. Frankly, if I had time, I'd happily spend time on that and learn more about it. I don't, though. Claude knows and does the guessing and checking. So I provide the concept and it translates into a horrible soup of yaml. Clearly I'm able to press forward with ignorance, which is dangerous. There's a real risk that I'll wind up with the kinds of unhealthy work that worries the author of the tweet.
Extinctions followed homo sapiens across the planet millennia before the emergence of the technologies that you seem to think make the world 'complicated.' The Greek work biblos, for book, derives from the name of the region of the Levant (Bublos) that produces much of the best paper in the ancient world, until people denuded it, turning it into a desert. Iran and Afghanistan were green when the Hittites and Babylonians were in charge, if I remember correctly.
Mostly I agree with overall perspective and tenor of the piece, but there's a profound absence of (historical) awareness, paired with a weird, presumptuous, sophomoric sanctimoniousness -- clearest in the strange insistence on using the word "we." If you've ever listened to recordings of sermons from Jamestown, you'll hear something similar: the breathless outrage and stupefaction at what "we" have become and what "we" do and "the world today." It's millenarianism and apocalypticism, and it's just goofy. It's the tone of a kid in his mid-teens who is worked up by his latest epiphany: he finally gets it and is wildly excited to make it clear, and he's performing it and acting it out for his parents, showing how serious he is -- and all the adults in the room know that he's on his way to figuring something out but doesn't grasp that he's trying on an idea and a personality to see how it feels. I hear the same cluelessness in this piece.
> Don't be curmudgeonly. Thoughtful criticism is fine, but please don't be rigidly or generically negative.
Could you explain how this applies? I don't mean to be difficult -- really. I appreciate that Hacker News is a great place for discussion, and I very much appreciate that's partly because of the work you're doing.
I'd just like to think that my criticism is thoughtful and is neither rigidly nor generically negative: I pointed out specific omissions and offered an analogy to explain the immaturity and cluelessness that I see in the piece -- not just that the claims are wrong but that the perspective is delivered so badly that it's difficul to take seriously. It wasn't meant to be unkind or a swipe; I didn't call names or sneer; it wasn't a generic tangent. It was the best way I could find to characterize the ways the piece's tone and content work together, undermining it and (almost certainly) rubbing people the wrong way.
> [the whole Jonestown (?) bit as a metaphor – at least a little disproportionate, no?]
> I hear the same cluelessness in this piece
I guess the issue is the number of negative terms and characterizations all crammed into one paragraph. It just seems to lay it on too heavy.
The writer is clearly wrestling with something and trying to process it, and the post has tapped into broader sentiment here, given the amount of front page time and discussion, so the dismissal seemed excessively scathing.
It may be that you underestimated how strongly your words come across to the reader, which is a common pitfall with online discussion forums; words often don't seem as harsh when we formulate them in our own mind as they do when read by others.
> The Greek work biblos, for book, derives from the name of the region of the Levant (Bublos) that produces much of the best paper in the ancient world, until people denuded it, turning it into a desert. Iran and Afghanistan were green when the Hittites and Babylonians were in charge, if I remember correctly.
I was fascinated by this so I looked it up, it's mostly inaccurate, but your larger point remains valid.
1) The Greeks did refer to ancient Lebanon as Byblos, because they bought their paper from the port. The paper was actually made in Egypt and imported there for resale though. They did, and still do, have big trees in Lebanon. They were famous for the cedars. Most of the ancient cedar is long gone, but its still green.
2) Iran and Afghanistan basically have the same climate now they did then. Desert then, desert now. You may be thinking of Iraq. Mesopotamia (Iraq) did destroy the fertile crescent by over irrigating it for too long and basically salting the earth.
I say this with respect and appreciation for your thoughtful framing, as I also feel for the author:
I'm not a young man, but I believe your this-has-always-been-the-way-ism, is equally clueless, in shared lineage with all the old-dog elders of past who've been helpless to stop what's happening, as the naive fools do the work of imagining it might be otherwise
Blindness goes both ways (a certain type from the end, as from the beginning), and truth is likely somewhere in the middle
In what way is understanding the historical context in which we live "blindness"?
Correcting someone who believes an old phenomenon is a new phenomenon, is not the same as giving up and saying we should do nothing about said phenomenon. In fact, understanding something is the first and most important step to changing it, especially a pattern or a habit.
I get into the this conversation a lot, when you point out the obvious historical context to "all this change", the response is always "Oh so you want to do nothing?" or "helpless to stop what's happening". That's not the implication of historical context. But it screams for a change in narrative, we aren't helpless, we live in the greatest time, and it can be even greater.
If we are to continue the march of civilization our algorithmic feed driven mania would just be just a blip. But if we give into the hysteria, I am afraid this is the beginning of the end. Our birth rate is dwindling because people are anxious [1], posts like this are not helping.
> enable entirely new modes of interaction. AI allows 3D content to be created in more efficient ways than ever before. A perfect and novel fit for spacial hardware and software, that natural habitat for 3D.
If we're certain, and it's clear, that the technology allows for "entirely new [anything]s," why aren't you listing any of them? What are they? Why do you think you can see these so clearly, whereas the teams and engineers actually working on the technology apparently didn't or don't?
How many 3d interfaces exist in the real world, in contrast to (essentially) 2d? Can you think of any computer interfaces where adding a third dimension resulted in a better experience?
> I believe Steve Jobs would have gone all in, to deliver the next big thing in software interfaces
Jobs was a proponent of skeuomorphism in interfaces, which is precisely not the "next big thing" in interfaces but rather "what we already know and are comfortable with."
> It is not at all clear which should be the output impedance of the amplifier as a function of frequency to ensure the least degradation of the sounds in comparison with the input signal.
> So it may happen that a vacuum-tube amplifier - loudspeaker system has actually a better overall fidelity than a typical audio amplifier that was designed to demonstrate a much higher fidelity on a resistive load (because thus the transfer function is easy to measure and correct, unlike the complete transfer function to sounds).
I don't know the electrical engineering at all, but I thought that this was a solved problem -- or that the degradation and mismatch were effectively negligible, well below the point of inaudibility.
I love this. I'm wondering: what skills did this project hone? What are you better at now than you were before you undertook it? Or was it just for fun?
reply