Ask HN: Boring but important tech no one is working on?

330 points by sremani ↗ HN
In the 2020s most old generation people are retiring and not only the replacement generations smaller but there is gap in generational knowledge transfer. What do you think is important tech out there in which are we are losing our collective knowledge and hard won wisdom?

463 comments

[ 2.2 ms ] story [ 189 ms ] thread
This may not be the type of answer you're looking for but near me insurance companies (AmFam, Sentry, etc) pay extremely well for the area and pay big money to purchase startups
What kinds of acquisitions? Like what product categories from your knowledge?
The problem I've found in this space is the data.

I and my colleagues have done consulting work for several insurance companies. I'd love to develop some sick claim processing/auditing software or data management solution for insurance, but it's next to impossible to get visibility into the data they're processing and where the improvement opportunities lie without working there or having an in.

I'm guessing the cycle is; work at an insurance company, figure out a solution to their problem, ask your boss if you can work on that solution, be told no, quit and start a startup to do it, sell it to that company, get bought by that company.

This sounds complicated but remember that there is little downside risk for the insurance company that buys you. If you don't make a good product, then you or your investors are out the cash, and they are out $0. This is probably 99% of cases. If you do make a good product, then they can just buy it later when the $ is worth less! (There is also a risk that a competitor buys you, I guess.)

> processing and where the improvement opportunities lie without working there or having an in.

This isn't institutionally crazy.

It's an old, old, story. Lots of people feel like they can come in from the outside to a complex domain, apply some "generic" techniques etc, and make changes with huge positive impact. They usually wrong, usually enough that many people feel safe just ignoring the possibility.

Chance of success is much higher by building capability internally for the techniques with people who already understand the domain well. This however runs into both internal politics and moribund institutions so can be a real challenge.

One of the big startup purchases that happened while I was there was of a company that used machine learning to estimate the value of items in a room, ostensibly to guess how much your furniture and electronics are worth. Insurance companies are already good at actuarial sciences so they don't need you to estimate how likely someone is to file a claim, they need you to do things they're not good at
My thoughts exactly. Knowledge transfer in manufacturing / industrial environments is something that I'm working on.

- Language models / NLP applications for processing large amount of technical text data (SOP, documentation, technical data, machine text logs, voice to text, video data processing for speeding up corrective action, training, onboarding and highlighting areas of improvement / bottlenecks), digitising documents and extracting failure reasons / equipment names / spare parts / processes involved and making associations between them for pareto analysis, better search or process improvement recommendations

- Recommending the next steps to fix something / remote intervention / do something etc. Lowering the expertise threshold required for technicians, electricians, mechanics or reliability engineers to be effective.

- Enabling operators to become data scientists by enabling to train AI models via their day to day activities / analysis. Building better UX in general and providing simple tools that even a toddler could use.

- Autonomous factory use-cases / supply chain automation.

Would love to discuss with people who find these things exciting

I would +1 this

I think the long term best thing we can be doing is documenting the how and why of building, designing, fixing, working with etc... everything

Many companies store huge amounts of documents, but every team does it in its own way. One template can quickly result in thousands of variations. If robust documentation principles are not used from the very beginning (checklists / reduction in free text, visual indications, etc), it will be a nightmare to make sense of that data afterwards. Also, there is no value in generating large amounts of text data unless you can easily scan it and retrieve the information of interest
What I wrote wasn't simply a poorly specified product requirement document ;) Instead it was a general idea.

Better and more granular documentation in a form that can be interpreted by humans and also machine readable would be a desirable outcome for any system. Especially true for systems in which the builders and operators are being replaced or EOLed

Better?

I would not just knowledge transfer, but knowledge organization. We have so many different ways to represent knowledge, but it is very hard to access it or know where to look.

I think better training and investigation into best practices of organizing knowledge would benefit all industries.

Definitely, transfer can only happen when knowledge is organised and understandable by a variety of stakeholders, with different backgrounds (education, languages spoken, years of expertise)
I'd agree with knowledge organization. It's either you have to root through academic texts or try and navigate the spammy internet with no really happy medium. It's almost like there's a complete lack of quality middle ground information. It's either total SEO garbage or very low quality entry level information or incredibly specific/dense academic content and the middle ground is missing.
Back when I was in grad school, PhD theses were indispensable for actually learning my way around a topic.

The academic literature itself (even review papers) was way too terse and (I believe, semi-intentionally) obfuscatory.

I would assume that it hasn't changed much since you were in grad school.
Former librarian here, now at a tech co. This is exactly the domain of information science, and its a salutary tale in two industries talking past one another. Librarians have deep training in the science, philosophy and psychology of information storage and retrieval. Most of the time you think of things like Dewey numbers on library books buts its much more than that. At the dawn of the second internet age (circa 1991, think gopher, WAIS, Archie and a nascent thing called the web) there was a boatload of discussion around what this Internet thing would mean for information.

Then, the tech bros arrived and after a few abortive attempts to catalog things for themselves (webrings, portals, yahoo) the industry collectively shrugged and decided to ignore the problem, assuming that a search engine would always be able to pluck your favorite needle out of the haystack.

Except that, today, it cant. Intranets are essentially corporate graveyards of content. The public web is a webring of 7 or 8 megasites that vacuum up all searches and make it all but impossible to break out of their domain. That article you read in 2005 about XYZ? Forget it, you're never finding that with Google.

I'm curious what your thoughts about Cory Doctorow's Metacrap [1] essay which I think summarized a lot of the problems with the semantic, informational organization approaches of the early-mid web. Are you also familiar with research in informational sciences these days?

[1]: https://people.well.com/user/doctorow/metacrap.htm

Do you have any pointers to things worth reading? I've always sworn if I started a tech company one of the first 10 employees would be a librarian because it needs to be someone's job to organize the information, and you're right the automatic systems for doing it are horrible.

Search isn't enough because search doesn't help you if the philosophy behind how the information is organized doesn't make sense -- if what you need is spread around between 100 slack messages, emails, and unconnected unmaintained wiki pages. You need someone (or ideally a team) whose job it is to one one hand organize that information themselves, and on the other hand create a framework so it's easy for the non-librarians to put things in the right place.

But right now the majority of tech companies are like libraries without librarians, the patrons are just wandering around sticking the books on random shelves and wondering why nobody can ever find anything.

Organizing Knowledge by Patrick Lambe is what I started with.
Cool, thanks. Will check it out.
Yes, please tell us your recommendations for informational retrieval / taxonomy systems - what are the current best practices for the different mediums?
I understand why html is the way it is. I can recall looking at it very early on in its development.

But I think we need more structure to how we represent knowledge get to something like a semantic web.

If the knowledge was easier to parse, it would be easier to break out of the megasites.

Have you considered going into SEO? Combining SEO expertise w information science sounds like market dynamite. There is probably a 5k/month blog in simply applying information science concepts to SEO in practice, to say nothing of the consulting gigs, etc.
Lowering the expertise threshold required for technicians, electricians, mechanics or reliability engineers to be effective.

This is a really interesting application I hadn't considered before. Having lots of blue-collar family, helping new members of the trades upskill fast would take a considerable load off that workforce.

(comment deleted)
Would love to find out more about the lessons that you've learned in the process
There are hundred million dollar manufacturing companies out there tracking whole processes with pen and paper. Just sayin'.
*hundred billion dollar companies with a manufacturing component :)
start getting these gigs and boom, now you're a bonafide digital transformation consulting company
sounds awesome, have been working on several such project as well. would be great to share experiences
I was being a bit tongue-in-cheek because "digital transformation" has become an overused marketing term. But I think the core of it is valid -- a company has an inefficient process due to lack of technical expertise or whatever and you help them fix it.

There's multiple parts to this field of work: networking to find leads, doing discovery to understand a potential client's problem, formulating a technical solution, creating/negotiating contracts, and implementing the solution. My experience in this area was at a company that was large enough so that these pieces were split into different roles within our organization, and I was mostly on the tech/solution implementation side.

Interesting, that is my experience at Pfizer as well, where I lead projects end-to-end, from problem discovery to solution deployment and I pretty much did everything, from talking to coding
Sometimes they have in-house teams that are slowly switching things over. Are there any good blogs/articles/books on starting or running such a consultancy?
Yea, that's where the difference between contractor vs. consultant becomes a little blurry. No idea on books though, sorry.
I'm a partner in a factory and I believe this is an incredibly important area, and the requirements are fairly different than normal "just put it on Confluence" workplaces, in a way that most tech people don't understand and usually completely miss the mark when they're doing product dev.

- Your team is out on the floor. Their hands have grease on them. Using tablets sounds great until you're trying to use it with a glove on it, or your hands are dirty, and it's hard to get grease off tablets. But they need the info out on the floor. Also, it can be noisy on the floor.

- The team tends to be very visual. They don't like tapping on computers a lot. Literacy ranges from pretty good to kinda OK. Sometimes they refuse to get (or wear) reading glasses for whatever personal reason.

- They're working on proprietary hardware, but technicians with the right knowledge are not nearby to come in and look at it. You really need to be able to see the issues visually. Sometimes even hear them. AR might be interesting here. (I spend $10k to fly a tech out for a few days to look at a machine. The bigger issue is that I lose $10k a day from one machine being down, and a tech might not be available to fly out for a week.)

- Predictive maintenance. The fancy sensors and whatnot mostly don't work. Tech people try it in a clean, quiet office and it works, and they can raise money on it from clueless VCs, so money keeps getting set on fire with smart AI machine learning magic motor sensor companies.

- Preventative maintenance. How to schedule, how to verify it was done, how to check whether it revealed an issue that needs a follow-up. Getting people to do it, and verify it was done, can be a challenge, but there are huge returns to preventative maintenance (for example: checking gearbox oil levels, verifying lubrication line function, checking valve temperatures.)

- Diagnosing machine problems. Using prior problem documentation helps team members see most likely issues. But many of these people don't really want to sort through a database of prior similar issues because they "know" what the problem is. How do you provide this information to them in a way that feels more approachable to them?

I could go on forever. Manufacturing is an interesting environment because downtime is usually hundreds to tens of thousands of dollars of hard cost per hour, depending on the operation, and they will spend quite a bit of money to stop it from going down, but culturally there's a vast gulf between the white collar SF tech bros and what actually happens in manufacturing plants, so innovation tends to be more limited.

Predictive/preventive maintenance is actually a big thrust behind my current company, Dials.

HOAs, which we serve, are run by busy volunteers, yet expected to perform almost insane financial gymnastics, planning 30 years of major component replacement, e.g. common area roofs, piping, asphalt resurfacing. This involves (a) estimating each component's lifetime (total and remaining), (b) getting a cost estimate, and (c) coming up with a plan to spread paying for it out over however many years before it's needed, breaking that up between the units in the HOA, and collecting the funds, month after month.

People blame cultural issues ("people won't pay for maintenance") or "laziness" but the truth is, it's just too damn hard to do predictive/preventative without a very accurate inventory of what you have. You need to get all of this into a cloud environment, and then somehow expose it so that either internal staff or external vendors (more common) can see exactly what you have, bid on fixing it, and track status and work in a fine-grained way.

Our ultimate goal is doing the entire inventory automatically using computer vision (partner and I used to work in self-driving) and having enough data around that we can price and estimate everything accurately.

Nobody wants to pay for this as a standalone product so we just decided to build a payment collection product (for monthly dues), start with that, and build it up. It's going pretty well and we'd love to get more people on it. Email's in my profile in case you want to chat

Tracking stuff is hard. I wonder why QR codes won't work in this case, or something similar, or super basic otherwise / stickers or codes at first. Might be more annoying to generate and maintain them initially. CV could work really well to keep track of inspection steps as well, or to recommend what you should do next, and how to do it
We considered this, but it's another step. What I'm talking about is going to be hard and take a while, but feels like the "endgame" for how this is going to be done--automated, done with phones, no extra work.
This is really compelling!

How will you protect from incorrect estimates, insurance?

Spot on, many of these challenges are common across the board - from my father's plant to Pfizer and others I got the chance to work with. There is however a massive talent gap when it comes to high quality software / ML people in these industries as well. It's tough to get experts to generate quality data and 'recipes' for others to follow when their KPIs are not aligned. Maintenance and reliability don't seem to be sexy enough areas for management to invest in, especially if the value proposition is anecdotal at best. Would be great to chat about your approach for solving some of the above
It sounds really interesting to work in a manufacturing plant for a year or two in order to empathize with the industry and learn how to blend in software in a way that actually solves problems like you describe. Or, generally, to penetrate areas where technology solutions don't apply obviously. I wonder how you'd set that kind of arrangement up. If you could design a company around displacing a few cofounders for a few years where the product research is hands on, on the ground, doing the job, I bet there are many people who would be interested in this type of setup. I agree the software industry does way too much "solve for ourselves first" type of product development and it's really discouraging.
On-site visits, or contract work on the shop floor should be a good way in. Alternatively pro-bono work, part time or longer.
I think a key point that people fail to remember or marketing just oversells, is that there is no silver bullet for all the problems, especially for an industry that has so much history, precedent and inertia. People want to try and solve (and from the other side, want perfect full solutions) all problems at once, whilst in reality small improvements in key areas are probably the 80/20 that is needed to bring business value. I think continuous feedback and good "translators" would be key for any product in those industries. Manufacturing people are busy and will tell you what the surface level pain point is, but they don't have the time or maybe don't have the idea fully thought out on what the underlying problem/goal is.

After typing up all that, I realise that most of this is applicable to every industry.

I'm starting a new job doing exactly these things in order to reduce the carbon intensiveness of heavy industry, specifically cement production. I'm hyped because I think the technical challenges aren't too daunting, and the prize is huge.
I've started looking into CO2e reduction techniques as well. Would be great to discuss. Working with a client in the food space who is doing this just to learn more
These things are all on the radar of "innovation" types. I don't mean to say they're not interesting, but in the area of applied ML all this stuff is basically as mainstream as it comes (despite being unsupported by any actual research advances).
//Lowering the expertise threshold required for technicians, electricians, mechanics or reliability engineers to be effective.

Why is that important? Must every job be automated?

If it produces a better quality product/service with less volatility relative to cost, then yes.
Fewer and fewer people are interested in manufacturing jobs, especially the less glamorous ones. Large manufacturers are having a hard time using analytics and more advanced systems because of qualified labour shortages. I've spoken to manufacturers whose technicians can't even write or follow instructions correctly. Sometimes, sending 10 technicians to inspect an asset would results in 10 different opinions about possible issues / failures. All of these could lead to lower quality product and increased unscheduled downtimes, lower revenues etc etc. But, it is definitely important to still allow people to use their brains and come up with better options
Quality, reproducibility, and precision necessarily require removing the human.

If it's something "artisanal", that not necessarily true, but even then, intentional "mistakes" can be added [1]. Having humans for the sake of having humans isn't a charity that non-luxury businesses can support (à la Snow Crash). It'll have to be something that governments subsidize or enforce tyrannically.

1. https://www.forbes.com/sites/nadiaarumugam/2012/04/24/new-yo...

I volunteer for rural development organizations providing the kinds of services that suburbs would call dept of water or forestry. This point is very important to reliably onboard volunteers.

If we need excavation, our worst-case scenario is that that we need excavation by someone who also knows

I used to work in manufacturing as a process engineer (now a dev) and this is fascinating to learn about. Do you have any articles on your work or any use cases where I can learn more?
Machining.

No child left without a 5-axis milling machine!

Titan's of CNC and Haas both have great youtube channels addressing training because of exactly this.
>Titan's of CNC and Haas both have great youtube channels addressing training because of exactly this.

I would love to get into aerospace manufacturing. Any chance a code monkey can make the switch? What is the entry level pathway like?

Sort of related but I, a 28 year old tech guy, am going back to school part-time to get a BEng so I can get into more aerospace-heavy roles. It'll take 6 years but I'm currently happy with my work so am not in a rush. Currently I work in networking/search for a defense startup so it's tangentially related.
>Sort of related but I, a 28 year old tech guy, am going back to school part-time to get a BEng so I can get into more aerospace-heavy roles.

Best of luck to you friend. Absolutely worth it if you can hack it. I tried at 26 after a few years in the industry, and failed miserably. Protip: pay for a good math tutor.

It's the technician life for me, unfortunately.

aerospace -> code monkey seems to be the more common route to me but that might be sample bias
I've just noticed FlexCNC machines for the first time. It's a bridge CNC with a long bed, they start at $209,000, so I'm not getting one... but the idea of a CNC bed sufficiently long that you don't need pallet swapping, you just load jobs wherever the bridge isn't, seems like it could be quite productive.

They use a helical rack instead of a ball screw, which makes sense in 50 foot long machines. It's welded construction, so no huge castings, quite interesting.

We're going to need a lot of machining here in the US as the world deglobalizes.

(Civil) Nuclear technology.
Came here to say that. From what I understand, a lot of nuclear welders retired and because very few projects happened in the past couple of decades, their expertise was not transfered. It is a shame.
The 'advanced' nuclear industry has dozens of big players and is raising money like crazy. Everyone wants to mess around with exotic fuel and coolants. But no one cares about the real workhorse reactors anymore like ABWRs, APR-1400, etc. We should be more focused on those imho.
The ability for a single person to create a technology, digital or otherwise.
This is an interesting answer. What in your opinion are the main blockers to this ability?
Improving the core of the power grid to prepare for the near-doubling of power demand that will occur when everyone drives EVs, and also making the grid more maintainable, more robust, and easier to fix if damaged by anything from mundane causes to EMPs.

Some thoughts:

* Equipment to allow existing runs to be upgraded to higher voltages or even HVDC while being compatible on the other side with existing equipment and grid voltages. That way existing rights of way can carry far more power over existing wires.

* Smaller more modular cheaper substation equipment that can be slotted in and rapidly replaced, eventually to replace the mega-transformers and stuff that are very hard to physically ship places. Today there are substations with equipment so unwieldy that it would be physically challenging to ship replacement equipment.

* Lower cost methods of stringing new high power transmission lines to bring distant renewable energy to cities.

* Make undersea power lines as cheap (or nearly so) as undersea fiber allowing "global supergrid" systems to share renewable energy.

* Protection equipment or techniques to reduce damage from solar storms or EMP.

The power grid is becoming more and more central and essential to human life, and we are about to drop tons more demand on it via electrification of transport and HVAC (heat pumps). Yet it seems like there's not much innovation there. Right now if everyone comes home and plugs in their car it crashes the grid in many places, and this won't do.

Batteries are hot, but they're still too expensive to back up whole cities or regions. The need for batteries can be greatly reduced if the grid can be made bigger, wider, and at least an order of magnitude more reliable.

> Improving the core of the power grid to prepare for the near-doubling of power demand that will occur when everyone drives EVs,

The British National Grid says that already planned increases in capacity will cope with plenty to spare.

Your assertion is a very common one not backed by those in the industry.

Lots of people are working on this problem in different ways. But you're still right that it's a good answer to the question posed here!
Related question: this issue of power bill costs skyrocketing in europe and elsewhere, to where there is real risk of mass non-payment. Do we know what percentage of households being shut off would cause grid instability because of having to dump load from nuclear and other plants that can't easily be shut down?
I guess the simple way would be to cut off non-payers in smaller batches than would make problems for the grid.

But I guess there would also be industry willing to buy the excess immediately anyway.

Construction
I came here to say this. The newer generation are not craftsmen in the same ways as before. I'm a Software Engineer and it's the same mentality as somebody that is a Carpenter (my friend's dad is a carpenter that I have asked a lot of questions to).

The people that enjoy building and tinkering are more frequently doing this digitally. That is a big gap that isn't going to be filled as time goes on. "They don't make them like they used to" rings true.

I hired a carpenter a few years ago and we became friends / stay in touch. He had just started his own company in 2018 and the work he did for me he did with a “helper”. By 2020, he ran 5 crews and jumped in where needed to help keep deadlines and whatnot. In 2022, he said he hasn’t picked up a hammer in over a year other than training his crews of which he now has 12. All the growth has been entirely word of mouth, they all drive plain white construction vehicles without his company branding on them and he has never marketed his company at all.
What i mean is that construction as a whole is a completely overlooked market.

Working in that space since almost 5 years after product development in and for several industries for 20+ years.

It’s highly complex but the user group is an overlooked and diverse set of people that have to put up with less than crappy solutions for getting their jobs done - for the chaos that is construction.

So many opportunities and space for innovation that is not another food delivery service for meteopolitan areas but meaningful tools for people in dire need.

Love it and encourage everyone to have a look.

The trouble with construction is the Autodesk monopoly. They are kings of EEE.
There’s a whole lot more to it then what Autodesk offers…
https://en.wikipedia.org/wiki/Latham_Report

Latham identified industry inefficiencies, condemning existing industry practices as 'adversarial', 'ineffective', 'fragmented', 'incapable of delivering for its clients' and 'lacking respect for its employees'.

From 1994, almost two decades later and not much has changed.

Thank you.

And yes - not much to almost nothing has changed.

> Working in that space since almost 5 years after product development

Doing construction work, or software development for construction companies?

Work is being done in this area, but it doesn't move as fast as people would think. Milwaukee is doing a lot of electric/battery tool development that replaces/enhances gas tools, and there are new construction methods that take time to be adopted.

Two off-hand that I can think of are "cast and move" bridges and hollowcore.

Construction moves very slowly because of the permitting process and the equipment investments. You also have to deal with the realities - if you build a standard house via the methods everyone around you uses, you'll have no problem finding people to perform maintenance on it. If you build it differently, people may not know how to deal with it, even if the method is technically better.

Many "US builders" are stealing techniques from Europe and vice versa.

I'm currently working in this space and agree - very underserved in a lot of different ways. Commercial is more served because more money is involved, residential is underserved because of a wide range of reasons, and its a pretty real and large problem that will get worse with time without some real movement.
There’s process and digital tools on the one end and on the other the ugly truth that almost everyone building spaces - be it commercial or residential - will be fed up with the low wages they get for the treciourous work they have to do in most cases since the number of people in “western countries” willing to work in construction is rapidly falling.
Joe Armstrong gave a good talk on forgotten ideas in Computer Science.

https://www.youtube.com/watch?v=-I_jE0l7sYQ

Can't watch the video right now, but the slides (available at [0]) list these four "forgotten ideas":

1. Linda Tuple Spaces [1]

2. Flow based programming [2]

3. Xanadu [3]

4. Unix pipes [4]

[0] https://codesync.global/uploads/media/default/0001/01/de7dfa...

[1] https://en.wikipedia.org/wiki/Tuple_space

[2] https://en.wikipedia.org/wiki/Flow-based_programming

[3] https://en.wikipedia.org/wiki/Project_Xanadu

[4] https://en.wikipedia.org/wiki/Pipeline_(Unix)

Almost everyone who wants to fundamentally reinvent programming seems to come up with visual dataflow programming, which is strange to me because aren't those just function calls?
No... it's plumbing. Function calls are nested to-do lists. It's a different thing all together.

Flows all have queues.

Function calls (called messages) can go in a queue in the actor model.

I do think dataflow programming might be interesting for low-level if it compiled to hardware as well as C. There's an old project for that here but the actual webpage and docs are gone:

https://github.com/orcc/orcc

If I had to guess, I'd say food distribution and preservation. But that's just a guess.
Maybe IBM's Mainframe? I mean reliable hardware, etc.
Lisp, Smalltalk are the first family of important tech fading to oblivion due to time passing and corporate interest. Most people do not think them in such terms but they represent The Real Desktop tech, the user programmable flexible environment to live in to free the power of computing. Without it, like today, our civilization lost an immense opportunity. We need people that work to push desktop systems, single-applications flexible systems where anything is a function, easy to combine, change and extend as the user wish. Simple and powerful enough end users can use and change.

Domestic food production is another, far simpler, not that exiting, but still important: these days some children do not even know where came from something they eat. Without any catastrophic scenario any society need to know at least superficially anything essential to survive. This include for instance basic knowledge about tools for instance to grind and pack meat to make salami, tools to sterilize and store vacuumed foods etc new tools to modernize and made such process a pleasure to do.

Last but not least generic knowledge, generic tools. We have gazillions of hyper-specialist in any fields and veeeeery few able to see the big picture. We have gazillion of tools for doing a thing and only one. We need generic stuff. Standards not made like https://xkcd.com/927/ but made and updated to be useful and spread. This is the essence of most tech, including the generic desktops cited above, including solar panels with maaaany cells one after another, including bricks, simple thing we can use to made a wall, a house, a bridge, ... it's very hard to made anything in such domain, but it's tremendously useful once done. In the past we have seen some examples here and there, nowadays nobody care.

At least smalltalks legacy lives on in the _hyperscript project!
Not tech but I think millenials, gen Z, and younger have lost the know-how to make a dollar independently of mega corps. I never see kids going door to door mowing lawns anymore, for example.
You need a peaceful society for that. Also the State will fine you for not having a license to mow lawns. Something like that. Western world is a post-liberty world.
I disagree with this, I'm (unfortunately) swamped with short videos on how to 'make side hustles' doing just these sorts of things due to the cost of living crisis.
And almost all of these videos are scams, lies, MLMs, or too obvious to be helpful.
As a millennial here's my childhood experience with that:

I knocked on the door of an older lady (Silent Gen, or Greatest Gen) in my small town and negotiated to mow her lawn and the price.

An unemployed older Baby Boomer who lived halfway between us saw me and/or overheard our conversation.

As I was walking back to get my dad's push-mower & prepare it to run, this neighbor jumped on his riding lawnmower and quickly mowed the lady's lawn.

I arrived with my push mower in time to see the neighbor with the riding lawnmower accepting payment from the lady for my job.

This is not a metaphor; it's a true story. Make of it what you will...

Reminds me of when I started noticing adults in cars delivering newspaper instead of kids in bicycles around the time of the last recession
We get this in the UK too but often it's someone driving the car and a kid that gets out and actually delivers the paper?

We also have car washes that are either 100% automated or have 6 guys all cleaning one car. You can guess which is the most popular option.

I hear we have a productivity problem, no kidding.

Paper routes were pretty exploitative. The children, often around 12 years old, were/are treated as independent contractors who buy the papers and have to collect the money themselves. Deadbeats could avoid paying for some time, usually until a parent chewed them out. When you went on vacation you were responsible for finding your replacement and paying them.

Throw stranger danger into the mix and it's obvious why kids aren't doing paper routes anymore.

I'm pretty sure that's also because the number of households that get newspaper delivery now is so low, and so the density so low, that it's necessary to have a car to do it.
Did you give up mowing lawns forever based on that one experience? My stepson related a similar story once. He left out the part about he didn't return to mow until the next day. And the person who mowed was long gone when he arrived. Not implying you didn't make doing the job a priority. That, however, has been a common them for my stepson.
Non-tech DIY in general is becoming lost. There is so much knowledge available on YouTube but so many of my younger friends just pay people $hundreds to do very simple, quick things. I’m not talking about transmission replacement, but ridiculously simple things like fixing a leaking toilet and changing motor oil.
Changing motor oil always comes up but it is really worth paying someone a few bucks to do it for you.
Why? Especially if you have an SUV or higher car you can easily crawl under. Just unscrew the original plug, screw in a Fumoto valve, and an entire oil change is done in under 20 min, and then you just drop off the old oil at Autozone or whatever. While the oil is draining, you can change the air filters too.

Buy the Kirkland brand oil at Costco, and you do not have to trust whoever the mechanic shop hires to bust your car up or use low quality oil.

Option 1 requirements:

  - Equipment: wrenches, oil collection pan, oil container for transport to disposal, funnel?
  - Knowledge: Quality of oils, filters, Oil weights (5W20? 10W30?)
  - Willingness: To get dirty, to stain the driveway
  - Time: 30min if all goes well, then time to take waste oil to disposal
  - Trust: In self
  - Funds: $20 for oil and filter, assuming disposal is free
Option 2 requirements:

  - Time: 15min, plus a few minutes to drive to the shop
  - Trust: In "professionals", misplaced or not
  - Funds: $29.95
I change my own oil also, but I have no rational explanation for this choice.

I have the equipment, knowledge, and willingness to take the minor risk of mistake and deal with the consequences. $10 is not a fair wage for those prerequisites!

Of course I use better oil and filters than the shop would. Does this matter? Am I kidding myself? And someday I'll take my plastic milk jugs full of waste oil to a disposal place.

> I change my own oil also, but I have no rational explanation for this choice.

Of course you do. All the items you listed are a small, one time cost. Watch a YouTube video, buy the wrench, pan, and funnel ($50?). The “knowledge” of which oil to use takes 20 seconds to look up in the manual in the glove box of the car. And the disposal is pretty easy, you take the empty oil bottles, fill them back up with old oil, and dump it at Autozone.

And for this initial, let’s say 1 hour of work for procurement, you get to guarantee that the oil change on your car (which represents a significant portion of the average American’s wealth) is done properly, with zero risk of it going wrong.

Now, there is nothing objectively wrong with NOT wanting to do this and just preferring to pay someone else. But for a household with 2 ICE cars that they own for 10+ years, this is very little work for a lot of gain, and watching 1 less Netflix show is probably not going to hurt. The tools for for all of this also take a small space to store.

Which is why I do not consider it a “no brainer” to outsource oil changes.

> Of course I use better oil and filters than the shop would. Does this matter?

Why would it not if you intend on driving the car for 200k miles? My main concern is some minimum wage mechanic not doing the oil change properly. I know they are not getting paid enough to care, and if I have to stop and check to see if it was done properly, I might as well do it myself.

I've probably paid for 5 oil changes in my life. Many dozens done myself.

In those 5 (always while traveling cross country, and purposely done at a Sears or Firestone for national guarantees), I've had the following negative experiences:

  - 1 overtightened and stripped plug bolt, requiring three extra days in a hotel while the replacement pan (which they charged me for) was shipped
  - 1 case of badly greased up door, floor mats, and steering wheel
  - 2 improperly seated crush washers (or inadequately tightened) plug bolts, causing leaks in parking lots
In the latter two cases, I was hundreds of miles away before noticing, but the nearest franchise honored the guarantee.

So, I'm with you on the "trust" premium.

This can be simplified depending on your car. GM, for example, requires Dexos oil from any manufacturer. Pair that with an OEM filter and you have a 10,000 mile oil change.

The hard part is having the crush washer on hand and remembering to put it on. In my experience, shops have forgotten it too, used the old one, or have over torqued the plug screw. I think this is why many shops aspirate the oil instead.

I don't change my own oil for the same reason I don't mow my own lawn or haul my own garbage to the municipal dump.
Re. Changing motor oil example: I gave up doing this when my cars started having non-standard nuts on the oil sump. I had to pay the garage to fit a _headlight_ recently because the bulb was practically inaccessible without either specialist tools or 3 elbows :(
I wonder if that's partly due to people (both young people and the older people that would hire them) being uncertain if they can get away with it these days? There were a bunch of scandals about ten years or so ago about politicians running for office who had housekeepers and didn't pay social security. And kids sometimes get in trouble for running a lemonade stand without a food handler's card or business license or whatever.

There might also be a disconnect between what people think they should pay and what the cost of living is. In other words: maybe people don't do that sort of thing because they'd still be losing money even if they did it full time.

I mean we're in our 30's, even those millenials who do want to do yardwork as a career have hopefully moved beyond the "knocking on doors to mow lawns" stage.
All things related to mainframe, cics, cobol etc. It seems like it has been tried to replace it for decades. But it also seems like that replacement is not getting anywhere.
This is the most real answer so far. Knowledge is being lost in the field, very few want to learn the field, and it is needed in just about every company. Companies that existed in the 80s and have managed to migrate off of the mainframe are the very very small minority.

Pay is part of this, but I'd posit that most programmers would need to be paid a hefty premium over standard rates to be will to work on what they consider legacy or dying technology.

I found this to be an area that really doesn’t want motivated people, just entry level blank slate newcomers.

I tried to get into this field for four years and gave up. I’ve got 10+ years as a UNIX guy mind you. I participated in IBM master the mainframe every year to get real hands-on experience and practice. I did IBMs courses on Coursera for more practice. I read the red books to learn even more! I also took classes at an community college on IBM midranges too as a potential alternative to mainframes; I interned at a factory and worked with RPG-IV and AS/400 way back. Yet I didn’t get a single reply to a resume I sent out, not even a follow up email.

If you really look into this, IBM and their network of partner companies that do mainframes are really only looking for entry level folks out of college or 2 year schools. My guess is that companies have slashed their mainframe budgets so much they can’t afford non-entry level folks, and assumed experienced techs won’t take the pay cuts. It won’t be until the mass retirements that organizations will really panic I suppose…

It was at least fun learning the tech though, would love to had the opportunity to learn even more. Or even just have access to a system z to continue learning on…

I directly worked on the planning of a replacement deposits system for a major financial institution. The existing system was written in COBOL and the vendor did not want to continue any business, including maintenance support, on this old software. We were already able to extend the software ourselves without their help to add integrations or features.

Consultants came in and offered the following suggestions:

1) Write new software to replace. Estimate $3-400 million dollars.

2) Use a product that would transpile the COBOL to Java.

We chose option 3: offer a lump sump to license the software in perpetuity and maintain ourselves. Built a layer around it to better enable the systems of engagement use cases.

The problem with #1 and #2 is documenting all the use cases and validating them after changes are made. Can you even imagine the resulting Java code for #2? There are many edge case behaviors that were either purposeful or accidental, but now relied upon by downstream software.

To summarize, there's a huge price tag to replace these systems, it's an extremely high risk exercise, and the benefits to the business are at best incremental. It's easy to see why the financial sector won't bother until either regulators force them or they acquire another institution with a better system.

(comment deleted)
No single thing, but in general the gradually declining fraction of people capable of understanding complex systems "from top to bottom" concerns me.

I have seen far too many projects become locally-smart but globally-incoherent due to the gradual dissolution of systemic understanding over time.

That is by design. A person with complete knowlege is a liability. Workers should be replacable with minimal disruption. It's also less intellectual property to take with you when you leave.
That's why you don't want one person with top-to-bottom knowledge, but not why you don't want many people with top-to-bottom knowledge.
You don't want many people with top-to-bottom knowledge because those people command higher salaries, and because most jobs don't require top-to-bottom knowledge, meaning those higher salaries wouldn't translate into commensurate value being returned to the company.

Also people with top-to-bottom knowledge will tend to want to be promoted, and most won't (because employment is a pyramid, and also see the Gervais Principle) so you'll be left with a pool of frustrated, overpaid employees.

I think a decent rule of thumb might be to have three people with really deep holistic expertise. If there's so much context that it's hard for people to have and maintain that depth, then the organization is successful enough that it can afford to pay a premium for a few of them. And when one of them moves up or out, you have two more to pass on the knowledge to a new third. Maybe two or four or five would be fine too, but three seems to me like it might be about the optimum. But it's very risky to have either zero or one person who knows a system from top to bottom.
Physicist here. Hire more of us. Grokking complex systems, testing, and communicating our knowledge, are what we do.
Seriously. Did a physics bachelor's and taking a network science course in my master's right now. Also currently looking for a job, and it feels impossible to get people to take me, a generalist who can move up and down the problem stack with ease, seriously. I truly believe what Heinlein said; that specialization is for insects.
This is a puzzle to me, even late in my career. I think one thing is that as a company grows and matures, it gets more organized, and more formal about hiring. This makes it harder to hire misfits such as myself. HR has a hard time passing a misfit's resume along to a hiring manager. But also harder to get rid of us. ;-)

If bigger, more successful companies hire with an eye towards specialization, then the flip side is that there's a reward for workers who specialize. Also, many people like what they're doing, want to get good at it, and do more of it.

This is why, much as I love physics and don't regret studying it, I also won't tell anybody that it's a surefire career path.

“Locally smart but globally incoherent” of a great, pithy way to describe many systems/projects/organisations I’ve worked with.
I think there is some truth to the generational knowledge transfer because the tech industry is still young. That said, we should not expect the generation gap to play an outsized role in tech compared to manufacturing, engineering, etc.

Where the technology is mission critical, healthcare companies have worked for decades to train new developers on 1980’s tech.

In the same way that all cultural understanding can become extinct, we probably are losing some bits of wisdom as older generations age out. The problem is it’s not obvious what bits of wisdom are being lost, even to the aged.

Finally, given Moore’s law, many of the constraints which drove early decisions no longer apply. As long as we are on a hyper-growth curve in technology cost and adoption rates, we will suffer from a lack of wisdom, and need lots of extra energy to be spent on bad ideas in order to make progress.

Eventually, when growth rates slow down, we will have a more developed sense of what technological wisdom looks like in the long term. The good news is, that will happen on its own. The bad news is It’s a long way off.

Educational games. Hear me out. The way we teach us basically how we did it 500 years ago. This is stupid, boring and not scalable. We dont have enough teachers, attention span is short, education is costly. So we need something that scales, is fun and involves all types of media plus gamifies education. Think Skyrim or GTA meets MS Encarta
Interactive visualizations/simulations go a really far way towards helping students learn complex topics. Gamification is tough because you need a human to effectively react to a student's particular shortcomings and help them reframe problems.
I've been exploring Encarta 96 on a Windows 95 VM I've been curating for active nostalgia and I gotta say it's pretty awesome. I feel like we are missing out on some magic from those days. Even the intro screen still gives me goosebumps
I TA'ed at a tech camp a while back, and there was a 12 year old from Mexico city that had astonishingly good English. I asked him how he was so good at it and he just answered "Skyrim".

So yeah, kids soak up a lot of info through games.

I basically learnt most of my written English from games as a teenager. They didn't directly help as much with speaking, but hanging out in game chats or IRC for a ridiculous number of hours actually helped with that to an extent as well. Even though it's still written rather than spoken, it's still a more spontaneous form of communication.

I took English as a foreign language in school since I was 8 or 9, so I learnt some of the basics there. But if I was faced with a spontaneous conversation, no matter how simple, I easily got tongue-tied. (I was also shy and socially awkward in general, which didn't help.)

Spending those hours in chats is when I started getting more comfortable with spontaneous communication.

I'm not necessarily a huge fan of gamification in general, which might be a generational thing, but learning as a side product of something you're interested in can be a huge boost. It doesn't have to be games, though. Lots of people probably learnt a lot of English from music or books they were engrossed with.

Emmanuel Freund (Shadow - gaming pc in the cloud - founder) in France is working on PowerZ (https://powerz.tech/) and has raised more than $10M.
I credit a huge portion of my current career and general math and reading skills to having a couple CDs of https://en.wikipedia.org/wiki/JumpStart games early in my childhood. It's a pity I can't think of very many things that I would be comfortable pointing my own kids to and hope for the same effect.
I don't believe at all in gamification in education. The kick that inspires new scientists and engineers comes from the gratification of learning or figuring out something as nothing feels better than figuring out how/why something is happening and how to push it to it's edge cases. Even chasing grades is a perverse incentive IMHO.
PHDs are never going to be learning through games but I think at lower levels of education there's a ton of room to inspire a desire to learn something more deeply through games.
Kerbal Space Program seems like a game which both can both inspire newcomers and could also actually be used by PHDs to bounce around ideas. At the least it seems like some actual astrophysics PHDs rely on it for inspiration.
Think of it instead as designing tools that make learning fun, and enables that gratifying feeling you described. Lego, for example, has probably created generations of people with incredibly high levels of visual-spatial intuition who've all been inspired to go on to become designers and engineers of all kinds.

It doesn't have to occur within the walls of a school for something to be considered education.

Sure, I don't have an objection on that. In fact, anything up until college is actually distilled and idealised version of the ideas about how nature works. That's why people solve math, physics and other problems recreationally.

However I don't want to call it gamification because those are not designed in a game format in order to make you do something that you normally wouldn't want to do. Those are simplification or analogies to bring ideas within the grasp of the student so they can reach it with their current toolset but without mischaracterising it. So when a student calculates the trajectory of a ball in ideal condition, this is still a calculation of a balls trajectory and not something else.

>designing tools that make learning fun

While I don't wholly disagree, there's a lot of edtech out there premised on the notion that all subjects, even sometimes challenging ones, should be like playing a game for all people or you're doing it wrong. I submit that you're always going to have some percentage of students who pretty much hate math however you teach it and some percentage who don't want to read books even with relatively "fun" and easy options.

Can I find plenty of problems at all levels of education? Sure. Are there more engaging things we can do even in an environment that's largely not personalized? Of course. But everyone should have fun all the time is not really a realistic goal unless you just let kids do whatever they want.

Everyone should have fun all the time is not really a realistic goal, period. I don't think playing games is universally fun either. Games can be quite boring.

Games do often promote the concept of winning and losing, however.

Yep. Perfect example of this being the Lego Mindstorms platform. We had a "robotics club" at my private school which used the Mindstorms visual programming tools; there were 10 year olds learning about loops and variables in a way that was legitimately _fun_.
Depends on the domain. Mavis Beacon absolutely outclassed any of the "classwork" on typing I had in school.

edit: the 24 card games were really great too, though maybe only if you were already inclined towards math

For me it was flirting with girls on ICQ, and getting owned at StarCraft 1 before voip was a thing.

Nothing teaches typing better, in my experience.

Duolingo is doing it amazingly to learn a new language.
I speak 3 languages, attempted to learn the 4th one (German) on Duolingo and I find it very ineffective.

I recall having trouble learning English back in high school and then having a breakthrough. The breakthrough was, I figured out how to think in English when trying to debug my errors with the teacher who actually have lived in the USA. Once it clicked, I no longer had to memorise things but predict instead. My English is still not perfect but I can have fluent conversations and almost never need to look up words, I can figure out the meaning of idioms etc.

IMHO, the gamification in Duolingo is not geared towards making you figure out the mechanics behind the language. I think edge case exploration is way to go but the gamification usually revolves around memorising.

Duolingo is a pretty good tool for practice, but for trying to learn a language from scratch it's probably best to go through some YouTube videos for the pronunciation and some books first and consider it a kind of fancy flash card thing to do on the side.
This heavily depends on a course, and courses are different between pairs of languages.

E.g. learning Spanish from English on Duolingo is very natural, the course is built the way a baby learns a language, by practice and suggestion, with few small bits of grammar info strewn along the way. According to reports from my friends, learning Spanish from Russian on Duolingo is a completely different experience, all based on memorizing large amounts of rules before any practice even begins, and it works really poorly.

I assume that learning languages on Duolingo is all about building a chain of good courses based on languages you already know by the moment you consider another one. (Only partly joking.)

I agree. One skill is recognizing the meaning from a sentence already formed (what Duolingo puzzles consist of). A very different skill is coming up with the sentence in the target language without hints.

There is the company called Assimil that has small books based on this concept, where first you are filling in the blanks (not choose from preexisting options) and as the book progresses you have to fill in more and more of the sentence.

Educational games are not the same as "gamification" as an incentive structure, and I also think you're being very prescriptive of your own research style. I knew several academics in my PhD that thrived off being able to publish papers in better conferences, faster, etc than their peers. Also there's a pretty strong niche of puzzle game solvers (think the MIT Puzzle Hunts) that love thinking through problems in gamified ways.

Educational games are just games that also teach fun things along the way. I grew up playing the Carmen Sandiego games and it sparked a lifelong interest in history. When I got to college I took a bunch of history classes and even did a bit of undergrad research in it (well "history" is an incredibly broad topic and not something you research, but rather focusing on a topic that I was interested in) on the side. Think interactive fiction games exploring historical settings or works of literature. Games with some light math-based puzzling.

I don't think high level academic work can ever be taught through a game, but students feel the freedom to explore high level concepts when they are comfortable and confident with low level concepts, something that games can bring across better than the underpaid, overworked teacher in a lot of schools.

This comment reminds me of Status as a Service by Eugene Wei.

"Some people find status games distasteful. Despite this, everyone I know is engaged in multiple status games. Some people sneer at people hashtag spamming on Instagram, but then retweet praise on Twitter. Others roll their eyes at photo albums of expensive meals on Facebook but then submit research papers to prestigious journals in the hopes of being published. Parents show off photos of their children performances at recitals, people preen in the mirror while assessing their outfits, employees flex on their peers in meetings, entrepreneurs complain about 30 under 30 lists while wishing to be on them, reporters check the Techmeme leaderboards; life is nothing if not a nested series of status contests."

Some status games are better than others. ie when the game requires some common good to be performed, or the mastery of a useful skill. Some status games reward things that only drag us down
As an aside, when people point out virtue signalling (I'm taking it as an analogue of status games) without adding some analysis my first assumption is that they are virtue-signalling to put others down or make others seem shallow (or themselves deep). The cornerstone of western morality is Jesus and he was quite the "virtue signaller". Point being that the observation on it's own is uninteresting without saying something about the context or consequences of a behaviour. There aren't many things that are /purely/ virtue signals (or status games) and those that are, are quite obvious.
Jesus spoke about humility and dying for your beliefs, which he did. Martyrs don't need to signal their virtue (or social status), they do the virtous thing.

Status is a public display of desireability. It doesn't have much to do with virtue, but everyone wants it anyway.

Don't blend those two concepts, they are separate. People who make a public display and performance out of their virtues, 'virtue signallers' miss the point. You do it in humility before God.

People who don't show off their status, get little benefit out of it.

But what's "common good" and a "useful skill" itself is defined by a status game.
The selection of which status games to play is itself a status game.
It’s possible to do none of those things.
I do none of those things.
come on, nobody is a polymath.[1]

There is some really boring stuff to slog through on the journey of a formal education.

In school, people tend to choose subjects they are fascinated with, to the detriment of the other subjects.

I think some subjects like history either benefit from actual life experience, or need some additional storytelling to appreciante.

we really don't need the "wooden straightback chairs in schoolrooms" mentality

gamification might smack of spoonfeeding, but it also might allow a much richer learning experience too.

[1] statistically speaking

When I'm really in flow at work, it can almost feel like a game. Then I go back to yak shaving and jira.

There are (and could be better) environments that can isolate and focus on that. The combo of interface and specific challenge can really help communicate what's magical about engineering. Ideally that would help motivate people (versus just being a dopamine hit source).

Highly visual environments with rapid feedback seem to hit that button for me. The old bridge building game Pontifex was great, as was Besiege. I'd love to see more like that (there probably are, I've not played many games in years)

> The kick that inspires new scientists and engineers comes from the gratification of learning or figuring out something as nothing feels better than figuring out how/why something is happening and how to push it to it's edge cases.

This is also how Kerbal Space Program works, to name one example of an educational game.

Relevant XKCD: https://xkcd.com/1356/

There's a job I was considering recently that is space-related, I almost wanted to modify my resume with "I don't have professional experience with orbital mechanics, but I do have xxx hours in Kerbal Space Program on Steam."

I think the key is, nobody wants to be preached to. You need the game to naturally involve the skills you are trying to teach.
Sounds like ad-tech masquerading as ed-tech. There are a lot of ed-tech companies around. I don't know how many take this strategy, but I'm sure there is a few.
I agree. Anyone every play the educational mode in Assassins Creed Origins?

It basically took that game's massive, amazing and detailed map of ancient Egypt, removed all the combat and replaced it with what was essentially a huge virtual museum. You could go run around Egypt, go inside the temples & pyramids, explore the farms, cities and the Nile delta and listen to audio clips and view slideshows about actual Egyptian history.

It was legitimately really amazing and educational. It essentially piggybacked off of the colossal amount of work that goes into creating a AAA open world game's map and repurposed it to create something educational.

Would love to see more of that type of thing.

First time hearing about this! Not a big AC fan, but this is more than enough reason to snag it; is that a platform-specific mode?
It’s on PC, Xbox and PS at least, not sure about others. Really cool experience because it also includes pictures of artifacts and audio commentary from scholars. They did the same for Assassin’s Creed: Valhalla aka Viking Simulator.
It's also on Stadia, and also in the Odyssey (Ancient Greece) game.
Related tangent - I'd love to see more games (whether or not explicitly "educational") that are more collaborative and less focused on violence. I remember playing GTA and enjoying its then-new (to me, at least) open-world dynamics and imagining a more peaceful version or mode where you'd play the role of an EMT or field surgeon, running around helping people and saving lives or something.
SimCopter and Crazy Taxi are older titles that come to mind.
>> less focused on violence

> Crazy Taxi

You must have been a much better Crazy Taxi driver than my friends and I...

GTA 5 roleplaying servers are getting pretty popular now and do what you want. I haven't tried them yet but I've wanted what you describe and it might be fun to check them out.
You should check out Death Stranding! It has a unique asynchronous multiplayer model that rewards collaboration.

Players can place items and structures as they walk from place to place and you can see stuff placed by other players as you bring each in-game area "online". These items can be storage lockers, ladders and ropes to climb up difficult terrain, charging stations for your motorcycle, or even entire roads/highways stretching across a landscape. If you use an item from another player, you can give them likes. Note that you never see another player's character.

It's such a nice experience to come back to a river you bridged with a ladder and see hundreds of likes. There's not much combat in the game because it's generally easier to avoid confrontations, and the little combat that is present is heavily geared towards non-lethal weapons.

Anyone every play the educational mode in Assassins Creed Origins?

The sequels have this too.

(And, if you want to see what this stuff is like, there are YouTube videos that will show you; I remember seeing one by some historians reviewing the educational mode of the Greek game, Assassin's Creed:Odyssey).

(comment deleted)
Off-topic but a little tangential: There's a designer named Zander Whitehurst[0] on TikTok/Youtube Shorts teaching Figma in 30 to 60 second videos and I'm in love with the format. I do wonder whether it's effective, but I've learned quite a bit of Figma from him.

[0]: https://www.tiktok.com/@zander_whitehurst

Agree except for being "boring". I don't see educational games in anyway as boring. If you (reader, not iammjm) see it as boring, you are not the right person to build it.
Simply changing a game's locale to the target language you are trying to acquire is such an amazing immersion experience with the language -- and this wasn't even intended as an educational tool!

If playing Skyrim, you hear the target language constantly, utterances are subtitled so written and heard language can reinforce each other, and practically everything you look at is labeled in the target language. Best of all, you needn't feel so bad for spending all that time playing games!

The only reason why I speak English is because I spent my teen years flaming strangers on Slashdot, you insensitive clod.
I always like reading about, and comments from, ESL people who learned basically Internet English.

Subtly different to UK and USA English but comprehensible by all.

Subtly different how?
I'm pretty sure even a casual observer could tell I'm not a native speaker given a sufficiently large sample of my writings. GP is right, the correct thing to do is to just embrace that I'll never be as fluent, as natural in English as I am in Italian.

Learning English at a relatively young age allowed me to talk to people literally across the ocean. This very conversation we're having right now would have been impossible otherwise.

I'm not upset in the slightest. I count my blessings.

Internet English has different "regional variety" than real life, for example I have litterally never heard "grok" in real life but I read it here fairly often.
"Grok" is not seen very much on the Internet outside of hacker/sci-fi fan circles. And it feels like using it in SF circles is very much marking you as An Old - it comes from Heinlein's "Stranger In A Strange Land", which is not a book that has aged well.
The difference isn't all that subtle. Uk and usa english have regional differences in their english that goes beyond putting the 'u' in colour.

Emphasis on streams of thinking and constructing sentences in ways that make sense to locals, is different to the formal-ish and almost toneless communication we do online.

Internet communication flattens out cultural shortcuts and cultural deepening you can do with fellow speakers in the same nation and locale.

Everything has to be expanded out so that you can 'see all the moving parts', whereas local english can be a lot more intuitive and cultural.

A beowulf cluster of Natalie Portmans eating some hot grits, you say?
Guild Wars had the feature of allowing to set a secondary language. This would translate all game text from the main language into the secondary one in real time as long as you hold the given key down. This meant that at any point, you didn't need to fetch a dictionary or translate a word, wasting time and immersion. I'm so sad I haven't found any other game with this feature.

I'm trying to play games with another language, and while sometimes it's ok, other times you really need to know the language to understand how understand mechanics that are introduced into the game. City Skylines was a super tough game to play in a learning language, but it would have been a nice learning exercise if I had the above feature.

This is really neat feature, I wish all games have such functionality (especially RPGs!)
I'm trying something similar with Japanese via Ghost of Tsushima (finally started playing it this month). I've lived in Japan 10+ years, only speak conversational Japanese....so far I don't think I've absorbed any linguistic improvements from using the game. I might just need more time with this style of immersion.
Totally this.

I was never stressed out in terms of English language (I'm from Poland) because I've rarely (never) switched game languages to any other language than English.

When I was ~10 years old I've played through whole Final Fantasy 8 which I was obsessed at the time thanks to my older brother with a PAPER dictionary on the desk (I was translating stuff that I didn't understand) and I think that was the most effective way and time when I've learned most of the English language in my life.

Now I'm trying to encourage slowly my kids and I switch game languages they play from time to time.

That's a neat story! I've read that there is a developmental change that happens in the brain around age 11 that has the effect of "switching off" a child's ability to simply absorb a new language. To the extent that this is true, it is so great that -- from your experience -- videos games could provide enough of an immersive experience for you to rapidly and enduringly acquire a new language. Hope your kids take to it as well!
I work for an EdTech company called Prodigy Game ( https://www.prodigygame.com ) - we have Math and English adventure games that help students in grades 1 - 8 learn at school or at home. While I don't think this is a good substitute for teachers / is going to disrupt the entire education space, I think it is an excellent supplement to what is currently in place, and can really help kids want to learn outside of schools. I'd be happy to see more products grow in this space for sure!
Typing of the Dead did more for my touch typing than anything else.

I remember loads of educational games from my youth.

I would like to see a fun game to help me with differential equations. People still want to learn more advanced topics and I feel that not enough edtech exists for learning beyond the formative years, and even less for beyond high school. As an example, Khan Academy has received much praise, but I was less impressed to be honest.
It's not a crazy idea, but also not a panacea.

So not crazy, people have been having for centuries (i.e. the game part predates "computer game" part).

I don't want to say there can't be a real breakthrough that makes this a much bigger educational impact, but it's probably important to remember that as an area lot of smart people have tried this pretty hard (including your characterization of open-world + encarta) , and results are not that impressive, so far.

(comment deleted)
Maybe. Maybe not. I think the questions of "who pays", "who profits", and "who controls the material" matters a lot. So does learning theory (I don't think games will be the only way to learn). It is definitely a field we should pay close attention to if we care about learning.

For anybody interested in the space, I strongly recommend "What Videogames Have to Teach Us About Learning And Literacy" (James Paul Gee)

When I read this I can't help but think of the meme that goes: Why don't we just like, put nicotine in salad so everyone can get addicted to salad?
I was watching a Pixar movie and wondered, "what if these resources were spent in creating, say, scientifically accurate animations of biological processes for people who are learning chemistry, anatomy, genetics, etc." We all have seen small animations of processes, but some are inaccurate or simplified due to time and money constraints. I've dreamed about this since I was a kid.
On the topic of 'scientifically accurate animations of biological processes'. You may know of these videos, but I thought these might be worth sharing for others interested in this topic:

Here is a TEDTalk by Drew Berry: Animations of unseeable biology. (uploaded in 2012) https://youtu.be/WFCvkkDSfIU

Here is also a more recent video by Veritasium on YT that also talks about these animations. (uploaded in 2017) https://youtu.be/X_tYrnv_o6A

I did not. That's fantastic. Thank you.
Agree, education via a game makes lots of sense. Math is one of those that is perfectly suited since repetition helps to learn it.

There have been companies that specilized in educational games in the past yet they are no more. In the late nighties there were a few that were giants. If I remember well Software Toolsworks was one of them. It got acquired but now it's gone.

I suspect the reason we don't see more of them is that education in the US is mostly free so it make's little sense for parents to spend money on something they already have. And teachers don't want to introduce something that threatens their lively hood and way of working. It's just a guess on my part.

That's why you'd probably want to target the homeschool market if you had a startup making educational games. Homeschooling seems to be increasingly popular since the pandemic so that's a growing market and one that is probably not as dominated by very religious families as it used to be.
> And teachers don't want to introduce something that threatens their lively hood and way of working. It's just a guess on my part.

This is wholy untrue, my mother has several decades in elementary education and they would love something like that. The problem is that teachers and schools often don't have the autonomy to acquire these things, a teacher can't buy 20 copies of a game for students as even at 10 bucks that's 200 dollars. So at the budget level required to make games available you then have to go up to the district level. (Stop me if this next part sounds familiar) the District administration is pretty far removed from the students that will be using the product and the teachers using it to teach. Because of this technology companies are incentivized to build their product to the wishes of the district administrators rather than for the students and teachers using the product. This process is helped by the trips, gifts, and "favors" the sales people can use while pitching to the district level decision makers.

So the district goes with the product that checks off the most items on a big checklist, as well as if they are a bad administrator, gives them more control over the students and classrooms. Meanwhile the product that was actually purchased doesn't actually help the students, or facilitate learning because it turns out it is way easier to put together a "study" that shows "leading educators" (who were former district administrators) assure that it will increase X metric by Y%.

Sorry, had to get this off my chest.

I pitched such a thing to a department of education grant program and got rejected. I called it the Meta Educational Environment for Simulation and Gaming (MEESG) (long before FB thought about turning into Meta). The pitch was to create a core open source platform for which subject matter experts could create their own specialized training packages, tied in with a web interface for teachers/students to track progress, gamifying the educational process, which was the key to engaging the students (feedback on progress).

Oh well, I still like the idea though. One of the biggest benefits was that certain trainings are either very expensive just due to materials, etc, or are very dangerous (high-voltage electrician stuff for example), and doing it virtually would have a much lower risk and cost less.

That's basically what I am doing with Quill.org (I'm the founder / executive director). Quill is a platform for learning tools and games - we support six web applications that plug into a common, open source platform where teachers can assign activities to students and monitor results. At the moment, our applications are all focused on literacy, and they are more "tool-like" than "game-like". However, we use AI to assess writing, and the AI creates a game-like experience of writing different responses and getting different pieces of feedback (somewhat like a MUD RPG). We're creating the applications in-house now, and the API is not public, but the hope is to open it up in the future to allow others to launch tools and game on our platform. Quill is now serving more than six million students across the United States, about 12% of all K-12 students.
As an alternative, wouldn't it be better to invest into games that teach you something or sharpen your thinking? Playing fun games that elicited curiosity and boosted logical thinking had a lot to do with how I ended up doing what I do for a living (software engineering).
Short attention span is in large part because of computers, phones, etc. So putting ed games on them doesn't seem to address the core problem.

Our solution has been a no screen time policy. Education is strictly non-digital and a lot of real world, hands on learning and dialogue with real people. And our kid has loved books since infancy.

Does it have to be so strict? I'm not sure, but I do know that our society is inundated with screens designed to addict us and if I give in a little, they just beg for more. We'll introduce screens eventually because it's inevitable, but delaying that as long as possible especially during the formative years has worked well for us so far.

I agree. Screens are the problem, not the solution.
Time to remaster Logical journey of the zoombinis. Such an amazing game.
It's actually been remastered a couple times. Wikipedia says there were remakes in 2001 and 2015.

(Though of course, 2015 was seven years ago... old enough to remaster again?)

As a kid, the Encarta virtual 3D tours of historical places was extremely fun. I used to spend hours running around and exploring those.
Recently I got the game "The Logical Journey of the Zoombinis" for my kids to play on steam since I got young kids and I remember loving that game when I was younger.

The game is amazing, the way it combines fun, learning, and engaging kids without ever feeling like it's a lesson. Seriously my 5 year old now has a basic understanding of set theory because of it.

It feels like educational games peaked in the late 90's and early 2000's and everything since then has been a regression that spends too much time trying to explain things to students rather than letting them discover and explore for themselves.

I loved this as a kid! And also played it for a couple days as an adult recently ha. Even though I don't play much games anymore.

Some of the logic puzzles were actually still very tricky. Was very impressed.

Why does education have to be entertaining? Interactive, sure. But not entertaining due to attention spans dwindling.
More like attention spans for boring stuff were always short for humans. Making it more entertaining means people don’t have such a hard time remaining focused.
Short attention spans can be absorbed into poetry/philosophy and expanded by chasing the meaning of the tiny sentences.

With skill you can put a lot of information into a shorter song. Chain those short sentences together and you can educate and entertain across a rythym.

Just saying it's possible.

If there’s anyone from Meta who works on Oculus here:

I want to build an educational VR app for Quest, but where would it go in the store?

There’s no educational section in the store. So can I get any sales for an educational app?

Games already do this all the time. Hades is essentially a fantastic retelling of many Greek mythologies.
Lots of sibling comments have talked about games to learn logic, maths, etc.

To me, the less explored area is games (or computer assistance) that explicitly teach how to know and master your own emotions and how to get on with other people.

Edit: also games to teach basic household finance, critical thinking, and practical probability. (Academic probability teaching totally slides by unacademic people.)

Life skills, in short.

Agree in principle, though taking the human (i.e. teacher) out of the loop is a very dangerous proposition (consider the current wave of book-banning in Texas).

At one level, even the idea of studying computer games as literary texts is quite tricky, as they tend to be expensive, require expensive hardware, require teachers to have played them (would you expect a teacher to teach a book/film without having read/watched it?), and everyone's experience of it diverges.

At the other end, the idea of a massive game (like the "Young Lady's Illustrated Primer" from "The Diamond Age" where you learn everything as you explore the game world sounds very appealing, but is a massive undertaking even for a game company (you're talking about codify most of the curriculum), so would need massive state funding. Parents tend to want things done at school as they had to do them at school (even if those were crap), so you'd struggle to find schools actually willing to take this on board (and that's before the resistance you'd get from teachers).

(comment deleted)
Strongly disagree. Teaching and learning have been solved problems for a long time, for those that want them solved.

"attention span is short" -- b/c we have stopped rewarding its development. The answer is insisting people develop longer attention windows, not trying to stuff real learning into atrophied attention windows.

any good open source ones people are involved in?
Totally agree, learning geometry or linear algebra through AR/VR would be so easy and intuitive
I think graphic text books is also a super interesting field that is starting to be more heavily explored.

I did play many MS DOS learning games as a kid and i think they were helpful. (math blaster would be one that i remember the most)

I don't think everything could be a game, but we could definitely teach with newer technologies. Think of how Scratch programming can be used to teach problem solving or even math concepts.

I think accessibility will depend on what we can run in the browser. And what we can run in the browser will depend a lot on where JavaScript goes.

Shameless plug. I'm building this, a fun but curricular platform for games focused on mathematics grades 8 onwards. Please check https://pledu.co/trigo-shooter (no login needed). The game levels are designed to gradually expose you to one concept at a time. Currently live with 3 games on Trigonometry, Vectors, co-ordinate geometry. I am all ears for feedback!
The controls are quite inconvenient. A common type of controls for top-down games is WASD movement relative to camera + mouse to rotate and shoot. If you spam shooting button, it starts showing the same "not enough power" warning many times. For shooting it would be better to simultaneously play the animation and spawn the bullet, not waiting for the animation. The tutorial windows can be replaced with text on screen to not block movement (controls, level objective; power warning replaced with animation maybe)
I would add to that educational animations. For example, in biology/medicine good animation are trully helpful to understand better and form more memorable associations. Animations cannot replace textbooks but can complement them and make learning much more easy.
Will I need to buy 800 edupoints so I can unlock the organic chemistry module? Or some micro transaction like that?
Attention span is not a well-defined term, and according to most researcher there is not even any reason to believe there is only a singular value — the “kid who can’t sit on his ass” can play a video game/learn something from youtube for hours just fine, it all depends on the activity at hand.

So the “decreasing attention span” is a myth.

I agree with the idea mostly.

Games are limited in the sense that they teach frozen symbols and mechanics. Great for learning a large number of mechanics, engineering concepts and 'settled science'.

Getting your hands on broad concepts, blank paper and a relationship with a master/teacher/prof will always be necessary to teach higher concepts.

I will also note that youtube, khan acamedy, and comment sections have done a lot of education, intentionally or otherwise and do a decent job at conveying the master's perspective on the topic being discussed. Video is education that can be done at a low cost vocally and visually, rather than developing it into a game, one of the most expensive and difficult art forms.

Strongly agree. I would be happy to see the emergence of a few startups doing this.
I work on a gamification product. We have schools as customers, they made a game that engaged the students and in an a/b testing showed much better results over time.

They pay 0.1% of what other, non-education customers pay. Because that's all they could spare and allow.

Needless to say, the servers alone cost more than what they pay. Not to talk about the massive amount of support they consume.

So if you wonder why it's not widespread - that's why.

Ding ding ding… we have a winner. Scrolled all the way down waiting for someone else to mention the hard truth that they have even less money for educational software that they have for teacher’s supplies, which is so bad you’ll find teachers having to pay for supplies for their class. It’s nice to suggest the idea, but it’s not currently a sustainable business for deeper educational material due to support costs and the low ability of the customer to pay in the first place.
The reality of "technology in the classroom" for me was just a bunch of classmates goofing around and getting distracted with computers, whilst they could be reading a book combined with interacting with each other.
>attention span is short,

Fix that first. Because without that, education only serve to create decoys/rubber stamps/plants.

You might find https://www.synthesis.com/ interesting. They develop games in-house that teach kids problem-solving, collaboration skills, dealing with unknowns/dynamic conditions, etc. etc.
(comment deleted)
I'm not seeing a more fundamental issue in this thread; What are we trying to teach, and why? It's easy to subject people to studying math, history, etc. What if we wanted to teach appropriate social interaction? Moral analysis (NOT morals)? Critical thinking? Developing self knowledge?

The problem is that we need interactions with live humans to explore these subjects. Gaming in the metaverse won't do the job.

This seems key to me. We need to teach people how to learn autonomously. And provide them with quality tools to learn from.
You're thinking on the right path, but I'd suggest going wider than videogames.

I built a math program in Newark that worked stunningly well because it took the level assessment/lessons/testing from Khan and gamified it.

Not gamification in the electronic badge/"you did it!!!"/points sense, but contests between kids, with teams rooting each other on and parents helping run the show.

Education needs genuine praise and involvement to inspire children to put in the necessary effort. But videogaming the experience is a solution in search of a problem. We need less isolation and more coaching.

Railroads. Specifically freight rail.

There is stunningly little literature on any R&D in conventional freight rail transport, something I discovered when looking for any published research some months back.

High-speed rail, yes. Regular old freight, no.

Contrast this with autonomous and electrified trucking as alternatives, with which rail could offer considerable synergies.

I'd suspect that break-bulk, trainset assembly and disassembly, and routing might all offer opportunities.

But ... nada.

I suspect other modalities within the transport sector might be similar, notably ocean shipping.

I know that BNSF did wonders with modernizing how they dispatch trains (the old way is build a train until it's as long as it can be and then send it off, which meant you never could really predict when something would arrive) and they also are doing monitoring, etc:

https://www.bnsf.com/news-media/railtalk/safety/artificial-i...

How much of it is fluff I don't know.

Check out Shift5
Care to unpack that / explain what element(s) are addressed?
Shift5 is a company doing the R&D and documentation on tech concepts around railroads, specifically fleet monitoring and security.
I think this has much to do with certain parties controlling most of the dataflow for NA rail freight moves. Definitely a problem on the ocean and in the port terminals as well. We should chat!
I wonder if that is the case outside of the English-language world.
If anywhere, I'd suspect China.

That literature would be pretty opaque to me.

I think there's a number of little regional rail companies popping up now that the bigs have gotten to big to care about those customers; I bet they're reinventing where they have to and would possibly even be willing to fund some knowledge collection and sharing efforts.
I saw a talk by an engineer from Parallel Systems [0] and they seem to be doing some very cool stuff for the future of rail. Autonomous Electric rail cars and some other interesting things

[0] https://moveparallel.com/

Very strange product. The benefit of traditional rail freight is scale and cost over long distance. The benefit of truck freight is last mile delivery (no need for warehouse storage, can make multiple stops, etc). Their product seems worse than either except that it’s electric. Fortunately, electric trucks are well within reach.

Cool that they are building the tech though, maybe it will get bought out by an actual logistics/freight company that can find a use for it.

Isn't that because rail freight is pretty much a solved problem with little left to research? Each one of those you mentioned have been discussed, optimised and reorganised hundreds if not thousands of times across hundreds of countries during the last more than a century of rail history. The answers are there, just not everyone bothers to look and apply them. Autonomous electric trains exist. Scheduling optimisations to the second are well developed.

However there's R&D into cheap electrifying with battery railcars - which is IMHO a dumb idea outside of niche railroads, electrification pays for itself in the medium to long term, a battery bandaid doesn't get you far.

Ans there's also a kind of innovation from India - dedicated rail freight corridors.

There's a argument to be made for this. As of about 30 years ago, I had the opportunity to observe a freight yard at which the tech was 1920s -- 1930s vintage (electromechanical relays controlling switches). I suspect there are many such instances still in operation. To an extent, what works, works, but there are also limitations and sheer maintenance.

But at a bigger-picture view, as I read what's being discussed for trucking, I keep thinking that far greater energy efficiencies are attainable by a switch to rail, though a huge problem remains in the flexibility and time elements. There have been expedited / express long-distance rail initiatives, though those have been on the range of ~14 days for coast-to-coast transit in the US.

There is a very small number of Class 1 railroads in the US: Amtrak, BNSF, Canadian National, Canadian Pacific, CSX, Kansas City Southern, Norfolk Southern, and Union Pacific. One of those is passenger-only, and two are Canadian-owned. That leaves five US-owned freight carriers. As with the broadband industry, these tend to operate independent territories and routes.

Trucking is also highly subsidized via highway money, with gas/usage taxes covering less than half of roads..= Take that away or make trucks pay for wear and tear and maybe we would see more investment in rail. Long haul trucking almost shouldn't exist given how much cheaper and more efficient rail is.
I recently ran across a mention that this was a deliberate decision made in the 1950s to help further break the power of railroad monopolies.

The decision had merits at the time. Side-effects may be overwhelming those now, though I'll note that with only five US majors, centralisation within rail remains a concern.

A better way to do that would have been something akin to the "Rail Packages" of the EU, most notably the open access part. Infrastructure should be separate from operations, and access should be given to anyone at equal priority and fees. Monopolies are effectively broken, and that works for rail as well any other type of infrastructure monopoly - ISPs, power, etc.
I wish I could remember the source, as I believe it had more context.

May have been Tim Wu's The Master Switch, though that doesn't quite seem right.

(comment deleted)
(comment deleted)
Self driving train seems like a no brainer. I suspect that even with the tech 10 years ago it could have been solved. Where are they and why isn't someone working on it?

I suspect when something is this obvious, yet we don't see a product, it has to do with someone in the industry fighting change and making sure it does not happen.

Probably little savings in just outsourcing that one (small team of) humans to AIs, and you lose your custodian for negotiating out-of-the-box situations. Seems easy enough though.
Rail already has a tremendous labour advantage over trucking. The principle limitations seem to be net transit speed (freight moves on average at or below 30 mph net, and peaks at 79 mph in the US) and the sheer delay and confusion in arranging shipments.

Freight's worked well for high-volume bulk commodities (coal, grain, tanker cars). Intermodal ("piggyback" or containerised cargo) works quite well, but still sees delays compared to trucking, and generally is not viable for distances under 500 mi / 1000 km (within the US). Sub-car shipping (equivalent of less-than-truckload or LTL for trucking) is a nightmare for shippers.

Rail doesn't seem to be integrated into much consumer logistics. Yes, the net flows are slower, but it's far less energy intensive than truck-based shipping (or air cargo), and should be far more economic. The logistics and scheduling though seem to be a real concern.

Trackage consolidation's been an issue, there's been a net decline of trackage for most of the 20th century AFAIU. That especially includes suburban and urban centre transport.

Railyards are still large-area operations which are expensive in terms of locked-in real estate within urban regions. There are major crossing points and exchanges (including both ports and mid-line switching operations) which have been major choke-points.

The basic technology works. It's tremendously efficient. It actually is a largely un-sung success story in the US. And yet it seems it could be so much better.

On the topic of chokepoints and tradeoffs, the question of upgrading freight lines around Chicago is pretty fascinating.
Chicago (or more accurately, Lake Michegan) is a pretty massive transportation disruptor generally.

Highway trucking as I-90, I-94, and I-80 converge is hellacious.

I was referring to rail lines, but I imagine the mess is multimodal.
An old college friend went to work for CSX (I think), and we caught up after he'd worked there for a few years. He said that most of his job consisted of digging into old COBOL code to explain why a train was routed a certain way (e.g. Reno -> Phoenix -> LA -> SF instead of Reno -> SF), and that he rarely ever changed the code. They just made summary documents saying "there was rail congestion on the normal route".

I've always hoped that he was just yanking my chain and that they did...well, anything really.

This seems like something Prolog could be good for.
I had an opportunity to work at a company that developed one of the more prominent software suites for railways in America.

Unfortunately, the pay was too low and the company was having people come into the office in June of 2020.

Industrial automation and manufacturing:

- Need a dedicated degree (2 year as well as 4 year) in Manufacturing

- Trade schools and apprenticeship programs that do not sound like scam

- Automated material movement and handling

- Machine vision, automation of quality assurance

- Automated mold making and plastic injection molding optimization (still done by humans and takes 4 months)

- Factory software (ERP, MRP, MES, SPC, etc.)

If you're working on any of these, it would be great to chat
Something I've noticed in this space is that there exist "standards" which are intended to make systems compatible and useful and modern, but all of these standards are extremely poorly written and cling to outdated paradigms and weird non-standard protocols that aren't used anywhere except small subsets of the manufacturing world. Often you'll have to implement interfaces yourself because there's only one reference C library available that hasn't had any maintenance for years and won't even compile.

At risk of someone posting the xkcd on standards, I would love to see an full industry wide consortium to establish _modern and well designed_ standards, led by people with experience making modern networked services and systems. Specifically around industrial machine controls and data/telemetry collection. A few standards exist that are roughly xml-rpc-ish, but they don't even follow that much as a standard. OPC-UA is the latest "standard" to be making the rounds and is (IMO) doomed to fail due to its layers of complexity and unclear specifications.

Traditional draftsmanship (not really tech is it?), as in with a pen/pencil on a draftsman table, has completely disappeared.

My father is a (relatively) recently retired Architect, he ran a small but specialist firm of about 20 people. By the time he retired about 5 years ago there was no one in the office that had ever been trained in traditional draftsmanship, only himself. It’s a lost art.

There are now almost two generations of working architects, those that learnt 2D CAD in the 90s and early 00s who still think in 3d and translate it themselves to 2d in cad (a little like traditional draftsmanship). And those since who have only ever worked in 3d and used the tooling to “project” 2d elevations and sections.

I trained at university in the early 2000s in Industrial Design, we did some traditional drafting lesions (maybe three or four weeks). But then jumped straight to 3d CAD and never looked back. I suspect they don’t even do those lessons now, and in fact most kids have probably done some 3D cad at school, maybe only 10% had when I started.

Do you want to bring this back? CAD is a productivity multiplier. It's like composing a document in MS Word vs hand writing it.
Yes, architecture and design is an art form, the medium in which you work influences the design process and the output you create. I’m not saying all architects and designers should return to the drafting table, but it is a shame that the knowledge of how to work like that is disappearing from the workplace.

Also, not everything always needs to be “efficient”, what’s wrong with working slowly if you can then create an even better result.

How does the slower process lead to a superior result, exactly? Are there designs that modern CAD systems can't achieve that the older process could? If so, perhaps it would be more practical to focus on closing that gap by improving our modern tooling.
Modern CAD systems tend to produce reproducible, very similar systems. Drafting produces individualized solutions. "Superior" depends on what you want.

(For housing, I want individualized so badly. If I see one more SillyValley SWE storehouse - excuse me, "luxury housing" - I'm going to vomit. They're all exactly the same. I'm fairly certain the firms involved have traded macros or something)

I'm not sure you can close that gap. I've done both drafting and CAD design (amateur level), and... you approach the space differently. Drafting almost forces you to have a plan, while CAD very much is "as you go".

Housing is super expensive in SFBA because it's all individual. In SF any housing project needs to make it through design meetings where all your neighbors make arbitrary aesthetic complaints about how it looks and how it casts shadows on things.

To get housing back to the rates it was built (and the price you could buy it at) in the 60s it needs to be a hundred times faster and more factory manufactured, not less.

Luckily, if you wait a hundred years people will like anything. That's the only reason people like mass-produced Eichlers, SF Victorians, and NYC brownstones.

I mean, I know the Bay Area fails to understand that, but you scale up housing by building more multi-level housing, not faster building of single homes.

And what multi-level housing there is being built is absolutely devoid of any creativity already, it can't really get more factory-like.

I’m struggling with this thread.

I’m not sure what’s superior in the old approach.

You’re able to perform much more robust and detailed calculations using computers compared to hand. You’re able to iterate more rapidly. The only thing I can think of is that the designer or engineer has more time to ruminate on their design. Or there is no artistry in a computer generated print.

It’s probably the bias in my training (mechanical engineering tech and not architecture or industrial design) and what I’m trying to accomplish through drafting, but I don’t get this romanticism for hand drafting.

My program, in the mid-2010s, had 13 weeks of hand drafting. The program still does that in the 2020s. My program emphasized sketching and drafting as a core component of the engineering design process. So hand drafting is still taught, and there may be sampling bias at play in the other comments. Or my hand drafting isn’t “real hand drafting.”

Sketching allows you to iterate through different designs early on. You then move into the detailed design which results in the manufacturing specs. For example, hole sizing and location, material thickness. We absolutely did all of these calculations by hand in our courses, so you need to be able to draw something out. You’d simplify and not worry about completely accurate proportions, though.

I guess before the rise of CAD systems, you’d hand over a bunch of rough design papers with the final dimensions for a drafter to prepare the final print.

However, with computers we now do the detailed design through computer aided engineering (CAE) software. This requires a 3D model of what you’re building. You run the simulation and find out that the thickness of the material isn’t suitable, so you change the thickness in the model and re-run the simulation.

This CAE process gives the engineer the 2D print for “free.” All the information needed to create the 2D print are embedded in the 3D model. You just pick which projections you want in 2D and you specify the dimensions and tolerance. Now that’s done and there is no need to draft it by hand.

A beautifully draughty schematic or plan is something to behold - with carefully considered weight of line, and perfectly executed detailing. This can absolutely achieved with 2D CAD draughting, and is indeed quicker than hand draughting, but with that productivity gain, thinking time is reduced. The knock on effect is timescales for project completion are lost, then everything becomes a kt of parts and looks the same. This has been worsened, IMHO especially in architecture, by BIM tools like Revit and ArchiCAD. There are an awful lot of "Revity" buildings going up...
My father has a background that's eerily similar to yours, right down to running a firm the same size, and he feels the same way you and yours does (so do I).

It truly is a lost art and there's a lot more hidden knowledge and reward that comes with knowing traditional drafting than it may appear on the surface. Patience, precision, a sense of intuitive aesthetics, and how to arrange a composition on paper to tell a story effectively ... thank you for sharing this comment, I'll have to tell him somebody else agrees. :)

I think this craft is well-documented in books, even if it's disappearing from common practice, no [1]?

In general, I've found books are an underrated source of technical information like this. The library of congress is probably doing a pretty good job of preserving the knowledge needed to bootstrap civilization.

For example, I have a few books on country wood-crafting that have enabled me to become pretty self-sufficient on a rural property. Even though there's no woodworking knowledge in my circle of friends or family.

YMMV depending on the craft, but the scope of knowledge documented in dead-tree mediums is both vast and deep.

[1]: For example, I found this with 40 seconds of effort. I know it's old, but that might contribute to it being more interesting: https://www.loc.gov/resource/gdcmassbookdig.easystepsinarch0...

My complaint about cad is the exact opposite. That those traditional techniques have too much influence on how people use cad and how cad software was designed. People use low level abstractions that leave far too much to interpretation.
I would love to hear more of your thoughts on that.

I could see that being the case with 2d focussed cad tools such as AutoCad, however “3d first” tools like Revit for architecture or SolidWorks for product design are so far removed form traditional drafting I don’t realy see any alignment.

I work in an infrastructure field with lots of engineering disciplines doing separate design. Designs are typically exchanged in AutoCAD even when a 3d tool are used to make them. Or AutoCAD is used to do design work in 3D. BIM exists but is just another data silo. A lot of the large engineering companies will try and push their own BIM solution which adds one more silo that no one else uses.

Everyone wants their own system and solution to be the single source of truth and that is a huge part of the problem. On a project of thousands of people there is never going to be a single source of truth. It's like expecting a globally distributed system to be in sync. Ultimately you have to chose between locking and eventual consistency and there is no other option. Having a single BIM system helps but is not a silver bullet. Because people will still get out of sync as other teams and tools are brought into play.

I agree. The purpose of plans is to transfer information about design intent for specific construction outcomes. Years ago, this was done by paper because there was not another efficient option. For the last few decades, this hasn’t been necessary, but we still produce paper products. Now roadway and earthwork folks tend to lean on digital models, but that’s after bidding and only for a handful of items. There’s “BIM” but nobody can actually tell you in concrete terms what that is.

I don’t know what we should be doing to deliver construction projects efficiently, but I’m pretty sure it’s not this.

To add to the agreement chain: completely concur that models > printed projections of these. Dropping to a paper based form will always result in information loss. The challenge is paper / pdf's are ubiquitous. Tools in the CAD space are built around vendor lock-in with non-trivial license costs. As a result there still needs to be a way to distribute that detail to all the people involved in delivery.
Tell your Dad to start making youtube videos- that seems like the best way to preserve that kind of knowledge.
Ha! Not sure that would happen, he’s practically a luddite.
Try Shapr3D on macOS and iOS and I think you'll be pleasantly surprised -- maybe even shocked -- at how amazing CAD software can be. Should be. The UX quality is exceptional and inspiring.

P.S. I have no connection to the product or company (based in Budapest I think).

Oh now that looks interesting, I haven’t come across that. Being based on Parasolid should make it a rock solid cad tool. Will have to give it a go!
A family relative of mine is a qualified draftsman and unfortunately had to pick up stacking shelves at the grocery store because he wasn't able to get enough work ad a draftsman to support his family.

I suspect an architect + engineer is all that's really needed these days, or perhaps even an architect that knows enough engineering, or an engineer that knows enough architecture.

There are shockingly few people maintaining the video encoding packages that most people rely on.

https://xkcd.com/2347/ makes me think the same is true of image software as well.

I'm pretty sure there's a timezone database that's maintained by two guys and literally everything depends on it to get timezones right.

When I was at Microsoft ~10 years ago there was a single guy who truly understood how tables in Word worked. I'm sure they've fixed it by now but back then if you wanted to touch the tables code in Word you had to talk to the tables guy to make sure you wouldn't break everything.

If I remember correctly there's a lot of low level networking code that's basically in this state. I think ntp is maintained by a single person. Curl was written by a single person, not sure who maintains it.

Tons of stuff in the geospatial world is owned and maintained by surprisingly few people like proj and gdal. I'm almost certain I saw something about all GPS code in the world relying on a package maintained by a single person.

I wonder if this is actually a more efficient way to run things than every company implementing timezones/tables/curl themselves for each of their needs. current situation isnt ideal but not horrible.
It often is the better way, because you have ONE implementation that everyone uses (even if it might be wrong at some point) vs multiple implementations that are all subtly different and never align.
It is, but also its kinda concerning that some CS faculty member does this as a hobby[1].

[1]: https://samueli.ucla.edu/time-zone-king-how-one-ucla-compute...

Still is better than some big corp distinguished engineer equivalent doing it in his free time.

The CS prof probably has tenure , a lot of freedom on what he works on, never ending supply of TAs and students and can choose when he retires .

afaict, they're a "senior lecturer" not a tenure track faculty member.
, When I was at Microsoft ~10 years ago there was a single guy who truly understood how tables in Word worked. I'm sure they've fixed it by now but back then if you wanted to touch the tables code in Word you had to talk to the tables guy to make sure you wouldn't break everything.*

Every business I've worked in over the past twenty years have two things in common - one person who knows a key piece of thetechnology, and there's a woeful lack of documentation around processes and understanding. There's often reasonably good code docs, but that only ever tells you what the code is trying to do; it doesn't tell you why the code exists in the first place, or what the code should do. If you ever need to check the code is correct there's rarely any resources apart from that one person. In the worst cases that person left a few years before.

I've become quite good at technical writing because I always end up establishing a project to fix the docs.

> it doesn't tell you why the code exists in the first place, or what the code should do

This is one of the worst things in most codebases I see. Today I was told to use function X to solve my problem. Checked: 0 docs on what it does and why/how; although used extensively in the codebase. Read existing code and try to infer it, every damned time.

There are some bugs in browsers that are decades old and they do not get fixed, precisely because of this IMO - no one knows part X of codebase anymore, learning it would take weeks, as it's probably written in idiomatic C++ from 2003 :) and there are other priorities (mostly shipping new stuff).
Supply chain/logistics. I've been simultaneously surprised and unsurprised by the relatively minuscule number of people interested in driving technology and change for this space forward compared to other spaces.
I'm actively working in this space right now. Working with a Fortune 100 to replatform their entire supply chain software. Currently handles all $15B of their US revenue.
Fascinating, would be great to chat about your thoughts on the future of the space
Very cool! Feel free to reach out!
Would you mind pinging me? Might have something interesting. username @ very common google email service.
I was going to say fulfillment technology, which is a subset of supply chain / logistics.
Maybe you’d be surprised at how many people at Amazon work on logistics.
The work they do behind the scenes is very impressive. An "AWS, but for supply chain" would be huge in my (biased) opinion, though I'm curious whether they would be motivated enough to share their progress with the rest of the world since it's been quite a competitive advantage.
In a sense they already do with “fulfillment by Amazon”. The only thing missing compared to UPS or FedEx is the retail locations and last mile pick up.
Maybe for driving technology specifically the space is small. But for supply chain management there are a number of players: SAP, Oracle, and Kinaxis to name three. There are more than a dozen though.
There are plenty SCM solutions and TMS products, agreed, however that isn't the change I'm describing. OTM, SAP S4/HANA, etc are heavily used and have been around for ages, yes -- even better systems exist. Problem is: the quality of data going into those systems has not improved much, the flow of that data is pretty poor, and the barriers to entry for basic visibility are high unless you're moving a lot of boxes.

Despite best efforts, even among the largest lines (Maersk + others via DCSA), EDI still reigns supreme in this space. Nothing inherently wrong with EDI itself but there are plenty problems with access. Port & terminal level visibility are in an even worse place.

Systems tuning, but it's an area that really got me to learn what I have over the years. I realize I'm a bit biased.

It's amazing the number of production issues I've fixed with simple sysctl parameters.

The Linux kernel defaults don't fit very well in this huge-cluster/scaling out world

One thing we're strongly missing is the kernel being able to notify that a sysctl parameter may help.

Sometimes it does things like "clock changed, is all hell breaking loose" but it could do much more when queues back up, etc.

Absolutely! So much of my 'bag of tricks' is having been around the block so many times that I can sniff these things out. It's a bad way for things to be!
In general, our industry lacks a sense of "mechanical sympathy"[0]. Whether it's understanding what knobs are tunable in an operating system or using streams for reading data to keep a consistent memory profile, we often don't dig deep. Ditto for folks understanding database systems.

I think in very big systems, you see folks dig deep, but it's not encouraged with smaller systems. I think there's a lot of inefficiently run services - not from a price perspective but from a "what could I really get out of 3 compute nodes and a well designed database schema."

[0]: https://mechanical-sympathy.blogspot.com/

I appreciate you introducing me to this concept -- it's something I've thought a lot about, but could never verbalize
I think about it a lot. A portion of my job has been either optimizing big systems or building very minimally invasive software that has to process a lot of data.

It’s akin to tuning a sports car. You can slap a big engine in a car, but the difference of a few seconds off of a lap is thousands of man-hours in optimization and tuning.