Rendered at 18:47:57 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
renehsz 12 hours ago [-]
It seems some people aren't happy about the fact that this course only glosses over each topic, teaching students only superficial knowledge instead of doing a deeper dive on one or a few.
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.
tekacs 11 hours ago [-]
A book recommended to us when applying to Trinity College, Cambridge was 'The New Turing Omnibus' by A.K. Dewdney [1].
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.
It says another user is currently borrowing the book. Is that really how archive.org works or what am I missing? Is there one other user borrowing it and now no one else can view it?
I LOVE this book. I worked through it in 2012 the summer before I went to college and learned so much that my intro CS class didn't really introduce any concepts I hadn't seen in one of the chapters.
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.
Witoso 6 hours ago [-]
Anyone have a recommendation like this for electronics? Breadth-first, short chapters, “know these all exist” rather than a deep textbook.
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.
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.
yepyoukno 4 hours ago [-]
Check out YouTube, there is a growing culture creating high quality shorts of DIY electronics.
This guy is great because he doesn’t talk, he just makes stuff:
Amazing book! I received it as a gift before going into CompSci in mid-90s and I kept it and referred back to it long after I got rid of all my uni textbooks.
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
gavin_gee 1 hours ago [-]
brilliant primers for asking good questions to AI
grilledchickenw 9 hours ago [-]
181 GBP?! Now that is ridiculous.
EagnaIonat 8 hours ago [-]
Looks like a typo or money laundering. Here is a cheaper one.
A gut feeling I've been having since the start of the LLM era is that horizontal knowledge (aka, knowing that stuff exists, can be done in certain way, being able to understand the details without knowing them intimately and being able to do it with a wide array of things) will start getting more and more important the further we go in, because if you know that stuff exist you can point an LLM at it or get it to go in a specific way that it wouldn't take otherwise. LLM have a very strong tendency to just go from point A to point B, and they're often limited to what you can prompt.
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
poopingatwork 5 hours ago [-]
It's all about how much crap you can fit in your head without going crazy.
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.
erwincoumans 4 hours ago [-]
You may want T knowledge: deep vertical in one area and shallow horizontal in broad fields: Jack of all trades, master of one.
oscillonoscope 4 hours ago [-]
T knowledge has been valuable for a while but I think it's moving more generalized than that. Something like triangle or sawtooth knowledge. In my career, it's been rare to see deep specialists adequately utilized when hired. Almost every time I see them in a company, they're assuaging the company's neuroses about direction more than needing to synthesize new knowledge.
gavin_gee 1 hours ago [-]
yeah I think T knowledge is dead. its all horizontal now
kapep 8 hours ago [-]
I'm so sure about that. LLMs seem to have that tendency when people ask them to solve a certain issue, but they are just as good for providing an overview of a topic - or even better, because details don't matter as much.
alex_smart 7 hours ago [-]
Isn’t the LLM already better at broadly knowing everything that exists than any human could ever hope to be?
PaulRobinson 6 hours ago [-]
I’d say a SOTA LLM today has median depth of expertise over many, many disciplines, and slightly better than median expertise in a few key areas. “Jack of all trades, master of none”, springs to mind, but I’d say in the last 6 months it’s got above the median in some key areas.
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.
qalmakka 7 hours ago [-]
yep, but it's not better at picking strong, long term decisions in my experience. The LLM most of the time is like a very knowledgeable employee that just does what it's told, and if the boss doesn't understand that you can do something, will often not ask for it to begin with.
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.
skeptic_ai 7 hours ago [-]
That’s why you just ask: do you think this is a good idea? Correct me if I’m wrong
wolttam 7 hours ago [-]
So all it takes is asking broad “hey LLM, is there any obviously better approach to what we’re doing?” Questions once in a while.
(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)
wtetzner 5 hours ago [-]
[dead]
epiccoleman 2 hours ago [-]
It's funny because - yes, it does seem like it should be that way. But I feel like a huge amount of the value I've gotten from LLMs has been specifically because I'm a generalist type and I can sort of quickly cut through the space of possibilities.
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.
LLMs currently are a great demonstration of the wide difference between knowing things and applying them in practice. Knowledge contained in LLMs is vast, but latent.
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.
pmontra 11 hours ago [-]
I fully agree and I add that this applies to every kind of job. I recently watched a plumber open and clean some pipes. I understood (at least at a superficial level) what he was doing and I've been able to complete the cleaning myself (he couldn't for a number of reasons that don't matter here.) Of course I've been clumsy and I took N times as long as he did but I finally got water again in my kitchen. Without watching him maybe I wouldn't have trusted myself to open some connections or I would have taken hours or days to reason about what was going on and probably do a lot of useless work.
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.
xbmcuser 11 hours ago [-]
This is the reason I think LLM are not being used to their full potential. A lot of busy office work can be automated if people knew they could get llm to make them scripts to do it. Where as currently a lot of people keep asking the llm to do things instead of asking the llm to build tools to do stuff.
sureglymop 11 hours ago [-]
I think it's interesting because LLMs (at least if used in the terminal) quickly create their own scripts to do a certain task.
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.
magicalhippo 10 hours ago [-]
> I think it's interesting because LLMs (at least if used in the terminal) quickly create their own scripts to do a certain task.
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.
It's also because the LLM training loop is too short-lived to give them feedback on how well repurposed their code can be. I imagine the labs are working on this to make it better, but it's fundamentally a much more expensive environment than training on few-script rollouts.
vjvjvjvjghv 2 hours ago [-]
I agree. It also makes things deterministic and testable again. I don't think I would let LLM do anything important directly. You never know what it will do tomorrow.
WillAdams 6 hours ago [-]
Can it though? I automated one aspect of my day job and promptly ran into upload limits.
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.
wtetzner 5 hours ago [-]
I don't get it. Were you uploading files to a server to process them? Couldn't you run a local process to do it (which is, I assume, how the LLM got around the upload issues)?
WillAdams 3 hours ago [-]
When uploading to a server (Microsoft's via Copilot) I ran into upload limits, ~80--100 files per day last I was able to try.
When running locally, it took the LLM 405 seconds to process two files.
kaashif 3 hours ago [-]
To answer the question, we need to know something, anything about the tasks you automated.
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.
WillAdams 1 hours ago [-]
The task is to rename scanned files of a check overlaid on one or more invoices using the amount of the check and its date and a couple of the numbers from one fo the invoices using the prompt:
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.
IX-103 2 hours ago [-]
In most cases you should have the LLM write a program to process the files instead of feeding them to the LLM itself.
CobrastanJorji 10 hours ago [-]
This is why "T-shaped" knowledge is so valuable. You need depth in a given area. You should absolutely be an expert at the thing that you're most immediately doing most of the time. But you should also have a great breadth of knowledge, especially within your field, and the reason is so that you recognize situations where a particular solution to a problem you encounter may be relevant. That gives you the opportunity to learn more about that solution.
lordnacho 11 hours ago [-]
Indeed, this is what university actually IS. You won't be an expert in anything much at all when you graduate, but a broad survey will tell you how experts divide up the field, what concepts matter, and you will have a decent BS detector for your area of study.
ivanjermakov 4 hours ago [-]
Not just university, all education is balance between "broad and shallow" and "narrow and deep". Learning is a funnel for becoming the only person in the world knowing something while having all foundational knowledge underneath.
kqr 9 hours ago [-]
This is true, but it surprised me, and I think many others don't know it either.
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.
maccard 7 hours ago [-]
I did competitive programming, and was terrible at it. (I liked it though). A few years after I graduated I noticed that I had a “tendency” to map problems that we had down to known algorithms that are just not intuitive. Prefix sum comes up all the time, like probably once a week in my work, and the code is quicker and just as easy to write as the naive version.
Darmani 9 hours ago [-]
> It seems some people aren't happy about the fact that this course only glosses over each topic, teaching students only superficial knowledge instead of doing a deeper dive on one or a few.
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.
laidoffamazon 1 hours ago [-]
The reason why it’s superficial is most of the students are already experts at at least one of these topics by the time they take the class, usually even before they attend undergrad at all.
dofm 9 hours ago [-]
Indeed, this is the kind of course that would be useful on any CS degree, because it places the specialism in a wider context.
rq1 8 hours ago [-]
Yes. An index is always useful.
santoshalper 6 hours ago [-]
I don't know if people still use the term, but long ago, this was referred to as a "survey course".
teiferer 6 hours ago [-]
> But those subject matter experts typically gain their expertise on the job,
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.)
mihaaly 7 hours ago [-]
Even in the ancient ages of no AI long long time ago, the benefit of university education was not in learning specific methods and approaches for specific use but in learning to learn specific methods and approaches for a specific use, including learning what the specific use entails.
baxtr 7 hours ago [-]
In my experience people are disappointed whenever their expectations are not met.
Maybe the course promises more than it delivers. It can still be very valuable.
LoganDark 11 hours ago [-]
I'm struggling to find prose. I just see lists of buzzwords. Am I reading it wrong? I struggle to see how this teaches superficial knowledge.
teraflop 10 hours ago [-]
This website is a syllabus and summary of a course, not the course itself. The actual course involves hands-on time in a lab, building things.
zakqwy 10 hours ago [-]
I took and later TA'd this course. The weekly pages that guide lectures are only part of the story; the best way to see what is actually taught is to peruse the sites students make to document their projects. Past experience varies, but in most cases folks come in with minimal knowledge of e.g. electronics, CAD, machining, programming, etc., or are perhaps specialists in just one of the disciplines. HTM(A)A is about the delta between the start and end of the semester for each individual, and is both a life-changing course and a massive time commitment for all involved.
vladoski 3 hours ago [-]
do you think that it would be possible to share the lectures and recitation videos for this year's course when it will finish? I'll be glad to watch them :)
fxwin 9 hours ago [-]
> the best way to see what is actually taught is to peruse the sites students make to document their projects
are those linked/listed somewhere?
Galorious 9 hours ago [-]
On the site itself; people > students
cjaackie 17 minutes ago [-]
this link should be less buried !
fxwin 8 hours ago [-]
Thanks! I had only checked project presentation + prior years and assumed people -> teaching personnel
Animats 12 hours ago [-]
That's just one course, not a full load. The pace of that course is insane. It's not clear how much of what they throw at students is intended to be used. The syllabus is huge but the actual projects aren't that complicated. They make students make their own PC board, but that's not all that hard if you have all the right tools. Which they probably do.
It's trying to do it in the kitchen sink that's miserable.
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.
WillAdams 6 hours ago [-]
In a recent video on Adams Savage's "Tested" Youtube channel, he put forth that in 10 years or so, desktop CNC machines will be pervasive (ob. discl., I work for a company which makes and sells such machines).
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.
These days of course one could directly make it in any number of free or affordable programs.
fxwin 9 hours ago [-]
> This is an 18 credit class, requiring ∼18 hours per week for lecture, design, training, prototyping, and documentation.
It might be "just one course", but that's a lot of work!
Darmani 9 hours ago [-]
Students in this class measure their weeks by how much time they don't spend on it. 40 hours a week is considered slacking off in this class. Everyone I know who has taken it reports very high levels of fluency in all these techniques.
j2kun 2 hours ago [-]
Is anyone supposed to be able to use this website for anything meaningful? No proper slides, videos, articles, or anything else someone not taking this course could use.
dr_dshiv 11 hours ago [-]
Just back from Shenzhen. Realizing there is a big gap in higher education for how to understand at-scale industrial processes. From prototype to product, there is a lot of know-how. Anyone familiar with instructions along these lines?
That's funny, I was going to suggest Bunnie's book, but this is the new edition of that book (updated by Naomi Wu). Thanks for the link!
rjsw 5 hours ago [-]
Good to read some news about Naomi.
WillAdams 4 hours ago [-]
Agreed.
I went ahead and placed an order, since it seemed the least I could do.
sampullman 11 hours ago [-]
It's pretty hard to generalize the necessary knowledge in a way that's useful and actually teachable. You pretty much have to go through the process yourself a few times, and even then, different factories have different capabilities and procedures, so unless you're making the exact same thing you made before, there's always a lot to learn when bringing up a production line.
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
In my university we had a major manufacturing engineering. Even that would probably not cover all, and my exposure to it is a decade old, but it is still highly relevant to my career and would be a great extension to TFA.
o4c 12 hours ago [-]
Offtopic,there is a free course available on YT on precision machine design by Alex Slocum.
I think what ever you need to learn you need more then just watching tutorials, you need to practice and practice and practice until it becomes part of your autopilot skills then after that you will be able to become a someone who really knows that, what ever it was and you can work and sell it to people not have time to learn that specific skill.
jm_redwood 5 hours ago [-]
A year ago I described myself as having a centipede shaped skillet as opposed to a T-shaped one. Extreme breadth with specialization in many verticals, many overlapping "Ts" so to speak. For all their faults, LLMs make it easier to work this way and much more rewarding.
Yeah! Interesting to see how this course presents technical solutions to unnamed (or maybe irrelevant) problems. The question, why would you even need to make these things, is at least hard to find.
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"
Trlog 5 hours ago [-]
Huh, interesting. 2003 and earlier just show the students enrolled in the class.
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.
mlmonkey 1 hours ago [-]
People who just have a superficial knowledge of what Epstein did are the quickest to jump on guilt by association.
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.
fooker 8 hours ago [-]
Where can I find good video lectures on this topic?
mrnonchalant 11 hours ago [-]
How does one best navigate this? Work their way through the recitations as a start?
epsteingpt 11 hours ago [-]
This is, hands down, one of the best classes in the already exceptional MIT Media Lab.
Gershenfeld is a low key genius, and this curriculum has been taught every year well before AI.
haters gonna hate.
cedijec334 8 hours ago [-]
Hahaha yeah people gonna hate friends of Epstein… (just noticed the username)
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.
thomascountz 11 hours ago [-]
You have to begin somewhere.
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.
pastel8739 12 hours ago [-]
Making things poorly is a lot more interesting than not making them at all. And then you know who to ask when you need part of it done well.
gadsdengrappler 11 hours ago [-]
A bit harsh. I don’t think anyone is expecting they can waltz out of this course ready for a career in CAD. It is intended for students coming from a wide range of backgrounds including non-technical, so a whirlwind tour of the fundamentals of many different fab and design tools is perfect. Cross-disciplinary awareness and curiosity is rarely a bad thing.
mitchweaver 11 hours ago [-]
Harsh? Hardly.
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.
thomascountz 9 hours ago [-]
We don't have to agree, but I think your comment is a little harsh.
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?
Except it looks like these students are expected to create at least basic items. There's likely a lab component.
cedijec334 8 hours ago [-]
[dead]
haunter 9 hours ago [-]
> The use of lowercase throughout doesn't inspire much confidence either.
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
whatever1 12 hours ago [-]
I recently just prompted fable to make me an stl for a hobby of mine. You would be surprised by its success.
ryandetzel 12 hours ago [-]
Check out solidcraft. I spent the weekend with multiple codex sessions watching it build 3d prints for me. What's nice is I can use the tool with it to make my own changes like fusion 360 and so it's like a built in helper. Pretty slick
aaron695 12 hours ago [-]
[dead]
varjag 12 hours ago [-]
I did it with Codex once. The model didn't even bother to chamfer the edges but if you have no experience building physical stuff it still could come as a miracle.
JimmyBiscuit 9 hours ago [-]
Just tell it to do that then. Im flabbergasted at how great claude is with 3d modelling and cad work now. If you have a proper skill set up its like magic. At least for smaller projects
To be fair - they seem to focus on setting goals for learning, so that the student would have to learn it all on their own. It's like vibe coding. Teachers give prompts to students and students deliver on the learning work.
croes 12 hours ago [-]
Make a pencil
WillAdams 4 hours ago [-]
If you are seriously interested in doing so, there is a specialized tool for doing that:
The TechShop comparison rings true for me. I took a weekend laser cutting class at a makerspace near my old apartment back in 2019, and the first thing I made was a crooked little wooden box that barely closed. It still sits on my desk. The class moved fast, but having the machines right there made the theory stick in a way videos never did. I wish more cities still had a place like that.
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.
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://archive.org/details/newturingomnibus0000dewd/page/n2...
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.
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.
- 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:
https://youtu.be/_QAEILW-gCY
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
https://www.amazon.com/New-Turing-Omnibus-Sixty-Six-Excursio...
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
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.
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.
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.
(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)
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!)
[1]: https://epiccoleman.com/tripulator
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.
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:
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.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.
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.)
Maybe the course promises more than it delivers. It can still be very valuable.
are those linked/listed somewhere?
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.
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.
It might be "just one course", but that's a lot of work!
I went ahead and placed an order, since it seemed the least I could do.
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
Title: 2.75 FUNdaMENTALs of Design
https://www.youtube.com/watch?v=wgMq7yha9wU&list=PLksE8LDXGX...
"How to Make Something that Makes (almost) Anything"
https://fab.cba.mit.edu/classes/865.24/index.html
Nice!
https://news.ycombinator.com/item?id=44775830
https://1lib.sk/book/BvkA1X1JZL/the-secret-life-of-circuits-...
after reading this thank to the author in email.
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"
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.
Gershenfeld is a low key genius, and this curriculum has been taught every year well before AI.
haters gonna hate.
https://thetech.com/2026/02/02/epstein-mit-2026-documents
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.
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.
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
[2]: See: https://en.wikipedia.org/wiki/Maker_culture
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
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:
https://willadams.gitbook.io/design-into-3d
https://webperso.info.ucl.ac.be/~pvr/decon.html
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://bridgecitytools.com/products/pp-1-pencil-precision
https://www.youtube.com/watch?v=DNp_qGZ7c68