890 points | 6d ago | Discuss on Hacker News | Back to Radar
People dedicate entire careers to things like CAD — yet they somehow crammed it all into part of one lecture which seems to be little more than an enumeration of 50 different CAD tools.
The use of lowercase throughout doesn't inspire much confidence either.
This is like taking a "tour the crayon factory" episode of Mr. Rogers and making a college-level course out of it, but through the eyes of open source evangelist MIT-types.
In fact they would probably learn more from those old PBS shows meant for children, minus some right to repair propaganda thrown in for good measure.
I think it's harsh because you've qualified "poor" as a thing to be avoided without without understanding its purpose and history, or engaging with prior students' final projects and their motivations.[1] Even if you had, you seem to hold some beliefs about the Maker Movement and maker culture[2] (and higher education in general) that are incompatible with the course's aims. That makes your judgment harsh and unqualified.
I may be missing what an alternative would look like. Surely we shouldn't try to make digital manufacturing less accessible. Students pursuing degrees in industrial design, mechanical engineering, media arts, and other fields should get hands-on, cross-disciplinary exposure, shouldn't they? I'm struggling to imagine how we could take a different approach and achieve comparable outcomes. How would you ever learn something if you're not allowed to be a beginner?
[1]: Start here: https://pub.cba.mit.edu/863.26/site/file/prior.html
Going breadth-first instead of depth-first is a pedagogical choice and a good motivator which was less possible 20 year ago. Now, more than ever, being a generalist means you can actually make something end-to-end, and making doesn't mean you need a professional machinist, industrial designer, or firmware engineer. Finally, this is a practical, not a lecture-based course. Most of what's learned is in the lab where you're expected to work and problem-solve.
If you look at the student portfolios from over the years, I'd say this course has done a lot of good taking this approach.
I'm still surprised that people hang up on that in the big old 2026
And it's not a new thing either, like even writers in the typewriter age used it as a deliberate stylistic choice
University isn't about teaching you how to do a job, it's about exposing you to different ideas. This course isn't to teach someone who wants to be a CAD engineer how to use CAD tools, it's to confront students with a general overview of how things are made. It's making them a more well-rounded person.
I was around for the glory days of TechShop, took about thirteen of their classes, and they covered most of the techniques listed. But making actual stuff took much longer than the classes on machine operation.
I miss having access to CNC mills, laser cutters, and a water jet cutter. The computer-controlled stuff is easy to learn for a programmer. Welding requires endless practice and most people never get very good at it.
It might be "just one course", but that's a lot of work!
The software has certainly gotten much better --- one of my early projects had me drawing up vectors in Macromedia Freehand/MX, exporting them to make a font, and to import into a CAM program (MakerCAM, it was a Flash file, really), and coordinating the output from that with using the font in F-Engrave to V-carve a dial.
File and image at: https://community.carbide3d.com/t/locking-register-calipers-...
These days of course one could directly make it in any number of free or affordable programs.
My guess it that, today, the challenges to the desktop cnc are more of the operational issues, which could take lot longer to sort out than the toolpath programming, which an AI/CV could achieve on a shorter term, right?
I only have experience with the hobby-oriented machines, and have pretty much crashed-and-burned every time I've tried to do traditional interactive 3D CAD.
Not sure what AI/CV means in this context.
There are still makerspaces around many major cities. It sounds like you are in the Bay Area, Maker Nexus is still around. My understanding is that many of the staff/much of the community from TechShop moved to Maker Nexus.
Also not included are topics like "How to talk to customers" or "How to get customer feedback". Also not included, for example, "How to write a good story", "How to do crochet"
Title: 2.75 FUNdaMENTALs of Design
https://www.youtube.com/watch?v=wgMq7yha9wU&list=PLksE8LDXGX...
But let's not forget that an overview of what’s even possible is extremely valuable in its own right. When these students are later faced with a problem X, they'll remember that there was this thing called Y which might help. They won't know the details, but at least they'll know where to start looking.
This is especially important if you want to rise to managerial positions, and it might become more relevant now that AI agents are increasingly capable of implementing things: Solving problems is not all about how to do something, but also about knowing what can be done at all. Laypeople often have no idea what's possible, but engineers do, which lets them be creative with their tools.
All of this is not to downplay the value of true in-depth knowledge of a specific subject. But those subject matter experts typically gain their expertise on the job, or at a later stage in their academic life when they conduct proper research. An overview course like this one doesn't hurt even if many students will choose to specialize later on.
Going back to CS and IT, I found myself talking many times along the years with good software developers coming from a totally different education path (graphics, philosophy, agricultural science, etc) and they had blind spots in areas that I had been exposed to at least a bit at university. Examples: SQL, one of them knew perfectly well how the ORM of the day worked (some JS one 12+ years ago) but not why; networking, one of them knew only to enter the SSID and password and everything else was magic.
The issue is then, that those scripts are probably discarded after one use by the user, partly because the LLM doesn't make them generalized enough and focuses on solving the one task at that time. Partly because, when the issue next arises, it seems cheap to the user to just have the LLM do it again.
I think these models could be trained/fine tuned in a way to anticipate future issues or create general enough solutions but it would imply lowering their own use... so there might not be an incentive for that.
Let me just write a small Python script to decompile a section of this DLL so I can verify the constants for this function matches what the reference said.
With the images, I'll write a small Python script to compute the motion error using the Lucas-Kanade method[1].
Almost literally what Claude (Opus) casually has written as it chipped away at a couple of my problems.
[1]: https://en.wikipedia.org/wiki/Lucas%E2%80%93Kanade_method
When I found and configured a local LLM to handle files, it took 6 min. 45 seconds to process two files --- I need to handle almost 5,000 in Q4.
When running locally, it took the LLM 405 seconds to process two files.
If the task is filling in tax forms from a spreadsheet then yeah, you can ask an LLM to write a script for that.
If the task is proving thousands of math theorems or the task is translating specs into working software, then yeah, I guess you do need an LLM.
For each file which was uploaded create a move command in a .bat batch file which will rename the matching file using the current filename of the uploaded file and use the amount, and date of the check, the Invoice ID, PWS ID, and COMPANYNAME in all UPPERCASE run together with no spaces from the PWS population line using a filename like to:
$<check amount>_<Account ID>_<date of check as YYYY-MM-DD>_<Invoice ID>_<PWS ID>_<COMPANYNAME>.pdf
Do not explain how this is to be done, open each file, examine it, and create the appropriate command, returning the contents of a .bat file with each command on a separate line which may be run on the files in that directory to rename the files from the name as uploaded to the generated filename with the command on its own line.It's still to this day one of my favorite reads. By the time I read it, I knew about a lot of the possibilities within it already. But it does exactly this: it skims across the top of dozens of topics, and it makes you just aware enough of... all sorts of areas of CS.
If you're curious, you might dig deeper - at worst, you know these all exist.
[1]: https://www.amazon.com/New-Turing-Omnibus-Turning-Excursions...
https://www.amazon.com/New-Turing-Omnibus-Sixty-Six-Excursio...
P.S. As sibling noted, the new price in that Amazon link is completely insane. Getting it used for a few $/€/£ is well worth it though
I remember that the book uses a basic or python-like pseudo code and that their neural network implementation has a couple of very subtle bugs which prevent a direct translation to python. The surrounding explanation was correct, but the code wasn't, and so it (presumably unintentionally) forced me to really understand what was happening in order to get it working. I forget the bug, something like a dropped negative.
The information density of the book is just so incredible. Each chapter is like 6 pages long and somehow that's enough to go from zero to a working single-layer neural network, or boolean logic circuits, or error-correcting codes, or 50 other things. I remember it being almost addicting to work through.
- Practical Electronics for Inventors.
- Electronics All In One For Dummies.
Hope folks would chime in with more, I would love more recommendations here.
https://nostarch.com/secret-life-of-circuits#content
And there’s “Practical electronics for inventors”
https://www.mheducation.com/highered/mhp/product/practical-e...
But electronics is not that large of a field and understanding the mathematical models is kinda required to design stuff as components can be rather fragile and not have a lot of room to play around.
This guy is great because he doesn’t talk, he just makes stuff:
I got through most of Art of Electronics, and then found this, and wished I had just gone through this series the whole time. AoE is much more of a reference than a good learning-from-scratch resource.
https://archive.org/details/newturingomnibus0000dewd/page/n2...
The official ones are:
.gd, .pk, and .gl
https://en.wikipedia.org/wiki/Anna's_Archive for up to date ones.
Curious, what people think in the past 30 years should be added if they did a 2027 edition of the book? LLMs obviously. Virtual machines/hypervisors/confidentiality. Memory safety. Those are at least a few things that seem like potentially good additions.
After primary and secondary education – which were very broad – I had expected to learn in depth about something narrow when I went to university. A few years into a computer science programme, I was at first a little disappointed when I realised I had still only taken what could reasonably be called introductory courses in various sub-studies of computer science.
But as I got out into the workforce, I've noticed it's been useful to have tried a bit of everything and to have the vocabulary to talk about the basics in each of the sub-studies.
Then those people should be very happy to know that this course is the equivalent of multiple full time jobs. "I had 16 hours a week to spend on everything that's not this course" is something I've heard from multiple people who have taken it.
The "overview of what's possible" is instead instead provided by Intro to Making, the "baby" version of this course which I took.
This is basically the opposite of the last 50+ years where vertical knowledge (being a savant in a certain very specific field) was seen as the most important thing to have in a super specialised world.
This of course is just my perception, which needless to say, may be wildly off
for instance, if you use a shitty library X in a project, and you ask your frontier model "please change A -> B" the model will just do that, it will not tell you "well maybe you could also swap the whole thing with Y" because that's not its job - it's job it to change A to B.
LLMs can be quite good at that... sometimes, and maybe most times. Where they can fall short is when "good idea" is comprised of inter-locking conditional definitions and variable evaluation thresholds denominated in differently weighted currencies.
(I agree, they may not be good at making strong long-term decision still, but that may also be an issue of inadequate context a lot of the time)
Humans have a practically infinite "context window" (active memory), we make associations and vaguely remember that thing we heard about that might be useful here, let's explore this thread ("you're absolutely right!"). Agents, being powerful as they are, don't do that, they pattern-match on the task and use the obvious solution, but don't activate their latent knowledge until you point them in the right direction.
The obvious solution might be the best one in 80% of cases or if you don't really care to get the best results (and it's absolutely fine!), but if you do care in your particular case, you have to babysit your agents and really understand if they're doing what you had in mind.
The sweet spot is a true expert in a domain driving the LLM and keeping it honest, whether that be via agents/skills or reviewing work. In software dev, there’s huge tranches of work at junior/mid level where LLMs can just go, and anything where there is a clearly testable goal it can iterate on faster than a human expert, for example recent progresses in maths proofs.
But, I don’t think there’s an LLM out there that is an expert at everything. At least, not yet.
Just knowing what's possible, having been exposed to a bunch of ideas, libraries, tools, and, yes, reading too much Hacker News - seems like it's allowed me to ask good questions, propose interesting experiments, and drive my own knowledge forward faster with LLMs.
I don't know that it's a particularly unique skill, but from personal experience there seems to be a spectrum of ... creativity? Freedom of thought / expression? And people who are kinda stuck in their ways seem to get less out of LLMs than people who have some idea of what they want and even some direction in mind, but just need the details made frictionless.
I think I'm not doing the best job expressing what I'm on about, so here's an example - I have a little livecoding app[0], which I developed (vibe-coded) for a conference talk I gave in October[1]. It's been a lot of fun to play with and has formed a good surface for lots of different experiments that never made it into the codebase - just throwaway ideas, etc.
So when all the hype about Jev was going a month or two ago, my experiment was "can Jev sit in on the jam?" And what I did was hook it up to the app - not to the "musical context" information that other agents (agents in the Elixir sense, not really the 'llm agent' sense) in that app receive, but rather to the raw signal info that was coming out of the master audio node. And then Jev made a prediction about what note to play, and then that note was sent through WebMIDI to a VST on my machine so that Jev was almost like an outside "guy" soloing over the app's music.
(Jev was a pretty boring soloist, because playing the "most likely" note every time is not especially exciting, but that's beside the point here).
Maybe Claude / Codex could have come up with that idea without me - surely the models know about all of those different tools / options already. But I'd bet that if I hadn't known what to ask for, I'd have gotten something less fun.
So, idk. Does that mark me out as some kind of special generalist who will forever be ahead of the permanent underclass because he reads too much HN? Probably not. But it's at least a useful disposition in the LLM era.
[0]: https://github.com/epiccoleman/tripulator/ (AGPL, go nuts!)
I too have thought a lot about this because I've been able to get a lot of value out of lengthy, deep LLM sessions which bridge between domain exploration, research and ideation. A lot of the net value from the session isn't in the LLM's responses but in the notes I make as I'm spelunking different 'nodes in the knowledge graph' that I'm constructing, which is my own knowledge and understanding of a new, adjacent domain.
Working this way immediately felt comfortable to me because it mirrors the patterns I developed over a couple decades as a serial tech entrepreneur focused on identifying high-value customer problems which had been overlooked or insufficiently explored by others and then matching them to new technologies and processes to create a novel solution. Over the years, one pattern I found especially useful was hiring a university professor or grad student for a few hours of "Deep Q&A." My well-paid new academic friends often reported being exhausted afterward, with one describing it as "a high G-force, intensely non-linear Mr. Toad's Wild Ride" through his domain. Over years of doing this, I got pretty good at intuiting what I need to know, don't need to know and how to surface, then hone in on the boundaries and edge cases likely to yield whatever discontinuity I was intuitively sniffing out.
An entrepreneurship professor once asked to do a couple 'ride-alongs' to understand my "process" (though I've never called it that). His case study described it as "akin to an ADHD truffle pig unearthing raw, unrefined nuggets of economic opportunity". A key take-away was that nearly the entire value from the "process" was successively approximating down to the exact right "interesting question", far more than in any particular answer.
The nice thing about LLMs vs PhDs and grad students, is they're not unsettled when I hard-stop and reverse direction like an NFL running back or stall out retracing over the same seemingly barren patch until I 'pick up the signal' again and then go rocketing off in a different direction. To be clear, LLMs are not anywhere close to being a full replacement for the unique value my best human spelunking guides provide and I haven't yet seen any sign LLMs are improving in ways that will close that particular gap anytime soon (though they are improving in many other ways). I think the difference is the human domain experts have a degree of deep understanding with unbroken conceptual continuity from first principles all the way up the stack combined with a complete World Model. They can make insightful connections between concepts more than one or two degrees of separation apart (both vertically within a domain's 'abstraction stack' and horizontally across domains) in ways LLMs don't, except by widening their randomness thresholds which I've never found to be productive.
While an LLM's near-perfect recall of a vast web of statistically weighted, taxonomically-related facts along with incredible speed and infinite patience are super valuable, they're not a replacement for my domain expert human guides. For my needs, LLMs do provide around half the value of humans in this role, while being much cheaper with no vetting, introductions, appointments or breaks required. So, I'm using both in separate sessions - though an LLM able to fully track and engage both visually and verbally in a high-velocity, collaborative human brainstorm without being a 'third-wheel' would be an interesting design target for enterprising AI startup.
In short, I don't think you're wrong. There's real skill and unique value in being able to do, whatever it is I'm describing, consistently well - whether with a human or with an LLM.
But this:
>While an LLM's near-perfect recall of a vast web of statistically weighted, taxonomically-related facts along with incredible speed and infinite patience are super valuable, they're not a replacement for my domain expert human guides.
Really reads true to me - LLMs don't have a filter, they "know everything". So the value in the human is exactly the filter - they can kinda sieve through the mass of knowledge, weight what's important in context, and thereby extract something interesting from an LLM.
If I can get weird for a sec, there's something almost divinatory to LLMs when you're using them in this particular way - it's like a Tarot deck or I Ching or something, where you have this big collection of vague symbols (tarot) or "phrase" (i ching) that could almost mean anything. The human in that loop has to collate what subset of the symbols are interesting to the given situation. Without that filtering you just have this morass of information that doesn't necessarily get you anywhere, but when you're able to direct your attention to what's currently germane, you can get uncanny results!
(to be clear, I don't really think there's anything "magical" in any of these cases, but there's something to the metaphor!)
I don't think this is unique to LLMs but rather a serendipitous duplication of a well-studied idea in creativity and innovation research called bounded randomization or semi-arbitrary constraints. You can see the idea used as a creativity tool in various ideation frameworks like "Innovation Games." An LLM's "Temperature" essentially controls how a kind of randomization function is applied to its conceptual weight-space in a similar way to how 'Innovation Games' toolboxes include dice and even Tarot-inspired cards containing vague directional prompts. It's the same underlying concept as Rorschach's random inkblots.
> So the value in the human is exactly the filter
This is true not only in the case of my domain expert human guides but also in the case of a uniquely skilled human driving an LLM session. A good example is famous mathematician Terence Tao sharing his deep LLM sessions online which PhD mathematicians find immense value in. Obviously, those mathematicians can simply boot up the same LLM themselves, so the unique value isn't in the LLM's answers, it’s in how Terence steers the session: how he frames his prompts, interrogates the outputs, and curates which branches to pursue or prune.
Before I'd heard on HN about Terry sharing his sessions, I'd shared some of my 3 to 7 hour, 100+ round deep dive LLM sessions with a few experienced peers who asked to see them. The shocked responses I got from two out of the three were my first clue that maybe I was "driving the LLM" in a very different way than they'd ever considered. One described the vibe as more akin to Socratic dialog than Q&A. Early in the sessions, my prompts tended to be short and much of the value was in the LLM's longer answers. But as the sessions went on that gradually began to shift with my prompts growing more detailed and introducing new ideas, while the LLM's answers generally got shorter. By the end, as I curated away many of the possible directions and then developed the individual threads I'd teased out, almost all the interesting and valuable content was in my prompts, with final rounds focusing on weaving the remaining threads together into a new actionable thesis.
While I don't find even today's most advanced LLMs to yet be half-replacements for humans in filling this role, I do think LLMs can be incredibly valuable in other ways the best human 'guide / sparring partners' cannot. Unfortunately, the current single-threaded, chat-style interfaces are actually pretty terrible for these kinds of multi-branching, layered learning/ideation workflows.
While I'm getting significant value out of LLMs as a cognitive prosthetic that's part robot researcher, part fact-checker, part note-taker and part extrapolating mirror that restates and expands on my directional ideas (for better and worse), I've found struggling with LLM chat interfaces so annoying that a few months ago I spent several hours sketching out a feature set and UX that isn't a struggle to use in this way. Then the next week someone posted a 'Show HN' with their take on a tool for a similar workflow and several posters immediately chimed in with links to others - so we're definitely not alone in seeing the value. Show HN: https://news.ycombinator.com/item?id=49307700
I get a ton of value out of agentic dev stuff - arguably much more economic value from that side of things - but I get more enjoyment and edification from the same "sparring partner / smart journal / digital tarot deck" thing we're both circling around above. And the most interesting of those conversations definitely do have that branching nature you're referring to, and I'll find myself writing almost little essays back, trying to respond to various points, steer the discussion, etc. it feels like this kind of probing and winnowing of the space of the model, and the LLM is not really in the same role as it is when I'm using it for coding. It's this weird kind of jammy, vibey thing where you're kind of updating in real time and you're not necessarily looking for a specific answer or result but rather using the model as a sort of tool for thinking. It's really interesting! I can definitely see the value in being able to gain a little more control over the "graph" of the conversation.
Anyway - you ever feel like playing that game with a human, my email's in my profile. Always game for a good email pal. Thanks for taking the time to think out loud with me.
Either you fill it with vertical knowledge (which sucks because now AI wipes out a lot of specialists), or you fill it with horizontal knowledge.
Vertical knowledge is the deep silo of learning that comes from doing more advanced techniques to deal with increasingly difficult problems.
Horizontal knowledge is knowledge about process and workflow: where does the thing created by vertical knowledge fit on the bigger picture? What systems are integrated from left to right to keep you successful?
It may be that a successful developer is now a specialist in general technology.
Silos between specialists were always talked about being broken down but the truth is breaking down a silo means asking your specialists to do more work. You don't get to be a developer anymore, you must also take on operations for successful devops.
It may be that AI finally allows broken silos to be successful because AI removes a lot of friction that occurs when context switching between silos of specialists. And quality may mean more complex interactions and integrations, which increases difficulty of operations (notice a trend in outages lately?).
Then the question becomes: who is still creating quality with all these new tools?
Because at the end of the day, everyone wants maximum quality for the cheapest cost.
We will see more and more outages until these operations positions are filled (traditionally SREs helped with these issues but things have scaled too fast. Check GitHub growth for an example).
The value of efficiency goes up: who can create something and meet the market first without sacrificing quality.
Maybe the course promises more than it delivers. It can still be very valuable.
For some things, mainly practical applied skills, that is the case, but for a whole class of others it is definitely not. Theoretical foundations are one such area. Whatever you do not learn in college you will never pick up in the job. I have may colleagues who were weak on complexity theory or database theory or things like security fundamentals or number theory or even probability theory and they have never managed to bridge that knowledge gap compared to their colleagues who were strong in those things right out of the gate. Make sure to deep dive at uni, that is likely the only chance you get. (No, watching a bunch of youtube clips does not cut it.)
Agree 1000%.
Overviews are extremely valuable, especially given the synergistic complement of our new helpers that can work with us to drill down, learning/do, once we need more specifics.
All my internal wikis (my life is a wiki, my work and mission just the largest sub-wiki) operate on the basis that every topic page's first purpose is to provide links to narrower or peer information. And only secondarily, address itself in a topic-general way, not redundant with more specific pages.
Being able to navigate (mentally or via documents) at high speed, with broad domain visibility, skipping details until needed, saves so much time.
An intro/overview is gold. Overview = awareness, visibility, breadth, map.
In my experience the hardest parts are mechanical DFM and automating verification/QA. Choosing the right factory helps with that, and pretty much everything else
Gershenfeld is a low key genius, and this curriculum has been taught every year well before AI.
haters gonna hate.
I'd love to be able to audit at least parts of this.
The mass hysteria around the "Epstein files" is largely justified given how widespread abuse was, but many people are really trying to invoke guilt by association / guilt by proxy even when there is nothing particularly revolting about specific individuals mentioned in those files.
I would have thought HN would be more sensible in this discussion. Alas, newly-created throwaway accounts are already here trying to derail the discussion.
You have to remember: Epstein was always looking to expand his network and reaching out to any person of interest he could find. The fact that someone replied to his overture in and of iteself doesn't mean anything. He was like a fisherman, always casting lines into the water to see what bites. Once he got to know you better, he'd judge your value to him and your propensity to do illegal/immoral stuff and then blackmail you if you went along.
are those linked/listed somewhere?
https://academy.cba.mit.edu/classes/project_development/inde...
"How to Make Something that Makes (almost) Anything"
https://fab.cba.mit.edu/classes/865.24/index.html
Nice!
These days there is:
https://shapeokoenthusiasts.gitbook.io/shapeoko-cnc-a-to-z
and for the CAD/CAM side of things I have been working on:
It reminds me of the famous one about colors (which is a primer for computer people about the world of lawyers), "What Colour are your bits?":
https://ansuz.sooke.bc.ca/entry/23
I would love to read a more up to date one, this seems to be from ~1993, so 30 years ago. Has the situation changed, and if so in what direction?
https://1lib.sk/book/BvkA1X1JZL/the-secret-life-of-circuits-...
after reading this thank to the author in email.
I took this course in 2017 and it changed my life - I wasn't even an engineering major, just took it because some friend told me it was amazing. Especially for someone who had engaged with engineering in such a limited way, this course really opened my eyes to what was possible, and sparked a fascination for how things were built.
The pace is fast and you get a quick introduction to each topic, but the real learning happened in the investment you put in through lab time each week. You could just do the bare minimum but the folks who really excelled went deep at least on a few things, driven by their own interests and passion. By the end of the course we had to build a final project and the things people came up with and built by hand were unreal - full on CNC machines, gimbals, experimental art displays, etc.
This class made me reconsider my academic career at the time and pivot towards applied engineering and technology. I really miss the late nights in the lab with all the people in my cohort, and all the freedom we had back then to explore this wonderful space.
Comments are loaded live from Hacker News and are not stored by Mid or Real.