I'm super excited and bullish for manufacturing and I believe we are on the cusp of a manufacturing revolution. I believe we will get to a point in the next 100 years where many of our physical products are created at home, and instead of buying physical products, we will buy designs and "print" things at home. Distribution of physical goods will enjoy the same freedom the music in the 00's and video in the 10's enjoyed, with individuals being able to design and develop products and sell online without the logistics of distribution. Imagine being able to design a fork, spoon, and knife and sell it online for people to print out. Imagine being able to design a cup or a comb and offer it to people to print out.

3D Printers and CNCs are still marketed towards hobbyists and/or industry professionals similar to how computers were marketed in the early 80's. I believe in the next few years we will see the Personal Computer version of home manufacturing and a revolution will ensue.

I've never really understood the optimism about 3D printing of consumer goods. What goods exactly are you buying that can conceivably be 3D printed in plastic? Maybe some furniture and lighting fixtures, but that's about it, and that's something that ideally you would only buy a handful of times in your entire lifetime. Perhaps cutlery, plating and things like vases would be the same, but even those are extremely rare purchases.

So what are you left with that could possibly justify the cost of a 3D printer capable of printing a bed for you? Doodads and cheap plastic crap is better no consumed at all, rather than printing yourself some thingamajig, and is anyway already so cheap that getting it for free would hardly be an improvement.

I do see 3D printing as possibly a major advance for certain hobbies, where being able to create your own small parts for various uses can quickly justify even thousands of dollars of investment. But for someone who doesn't have any construction-like hobbies, I think there is really no reason for this optimism.

You can print other materials than just plastic. You are imagining today's 3D printers. OP spoke of a 100 year period. Computers are nothing like they were in their infancy, and there are many quotes of people claiming people would never need a computer at home.

Not that many, actually. You can't print most metals or glass (not without extreme temperatures), nor weaved materials. Maybe there are some ceramics that could be printed? You will also perhaps be able to print wood composites and a few other biological materials. Perhaps you may even print some electronics (though not something advanced like a microprocessor or FPGA).

If I'm missing somethign some examples of consumer goods you really think could be replaced by an advanced 3D printer that fits in an apartment or small-ish house (say, at maximum the size of a larger washing machine) would help a lot.

Again, you are comparing 100 years of exponential innovation to yesterday's technology. Yes, even today you can print many metals (gold, silver, titanium, aluminum, Tungsten, stainless steel, etc.) and you can print glass and fibers– let alone what will be possible in the future. And we aren't just talking 3D printing when discussing home manufacturing–there are reductive tools like CNCs and lasers as well.

You can walk through any Target or Walmart and see hundreds of items the could be made at home in the future.

https://www.3dprintingmedia.network/from-steel-to-metallic-g...

You make good points but I think stores like Amazon with fast shipping are likely to be more convenient for people. It seems the majority of consumers would rather buy something than do it themself

I wouldn't look towards your standard home as a test bed for this. I don't need to 3D print cutlery or even most things around the house. You just don't use that much stuff. The economics don't make sense for me to spend a large amount of cash to print household things.

Farms, ranches, and other remote businesses definitely have an opportunity for that though, because not only do they need a lot of every day things, they are also far away and sometimes things aren't in stock.

Sure, industry use is very exciting, and may well propel some 3D printer manufacturers to very high hills.

But the commenter I replied to was specifically talking about 3D printing in the home, so consumer goods, not industrial/business goods.

Perhaps we will walk to our local corner printing center.

It's a fun, satisfying hobby. You need to be willing to get your hands dirty with modeling, and be ready to engineer around the limitations of the technology, in order to unlock the potential as a tool for the home.

I print stuff for around the home all the time. It's great to be able to fix toys, closet doors, light fixtures, etc. My most recent print was a bunch of small stilts for a wooden playhouse we're building for our daughter. The playhouse will be on concrete in an uneven low spot, so I designed a piece that will take a 1/4" nut and bolt to allow the structure to be leveled, and keep the wood out of pooled water.

Over the years I've been working on a homemade force feedback steering wheel (for driving games). The gearbox is all 3D printed other than bearings, as well as a faux-wood dashboard. It's as performant as any commercially available force feedback wheel.

A 3D printer isn't going to evolve into some magical star trek replicator, though. It's a device for precisely making plastic objects within a bunch of constraints, or for precisely making resin into objects within other constraints.

Imagine genetically modified tree that grows into shape of a bed, seeds are customizable: pick a color, dimensions, additional features.

That would be ridiculously impractical.

Want a new bed? Just buy these seeds, plant them in a (at minimum) bed-sized pot, and don't forget to water, fertilize, keep in the right temperature range, and keep pests away for the next 1-2 years. Once it's fully grown, you'll just have to dig it out and allow a few more weeks until it stops growing new roots, a few more months until it dries into its final shape. Then, get ready to do some light sculpting to get rid of the root system, make good equal-sized flat feet, get rid of any branches that stick out. Now, buy some lacquer or paint to cover it thoroughly so it doesn't start rotting.

Hopefully it will not develop any disease or cancer in this time that could ruin its shape or kill it before it reaches the desired size.

They do this already with watermelons in Japan. They make square watermelons!

Mainly parts for other things I own, either to replace broken ones or to add a personal touch to my belongings

I don’t think 3D printers will ever get that ubiquitous. I think more likely modeling tooling will improve to the point that 3D prints can be done by 3rd parties for cheaper and higher quality than you can do yourself and those vendors can post the results to you. Sort of like PCB Way does with electronics.

I can imagine all sorts of things I’d like to design and make but unless I make it a full time job the amortized cost of the equipment will never make it worth while. I don’t think 3D printers will come down in price enough to change that.

> Sort of like PCB Way does with electronics.

PCBWay also[1] has 3D printing (including metals[2]), CNC machining, sheet metal fabrication and injection molding services[1].

[1]: https://www.pcbway.com/rapid-prototyping/

[2]: https://www.pcbway.com/rapid-prototyping/3d-printing/

Thanks, I did not know that, that is pretty awesome and it makes total sense for them to do that.

Right we don't need new things every day. Therefore we don't need our own 3D printer, we can share that with others. Let's order the "print" from Staples.

It depends on the resolution you're looking for but decent 3D printers are pretty cheap.

In the past there was a lot more local production. It switched to centralized production because that was cheaper and people valued cheap over custom. I doubt that will change anytime soon. Also a lot of the most useful materials are not amenable to 3d printing and CNC is non-trivial to operate because you have to deal with inventory, setup fixtures, cleanup parts, etc. Most people don't want to deal with the hassle and have other things they'd rather spend their time on. There are definitely a lot of cool things happening in this space, but I just don't see it being the revolutionary change other people do.

I think the shift to centralized production is not necessarily permanent.

Energy generation is decentralizing again. We can make more and more manufacturing feedstocks (metals, H2, CO2, biomolecules) using more modular processes that can also be decentralized.

Energy generation isn't really decentralizing in a meaningful way. Sure individuals can have their own solar panels or wind turbines (and for many people this makes sense), but the most efficient renewable projects are big centralized projects that take up a lot of land area. I believe it will always be that way for economy of scale reasons. This also means that large electricity consumers who can position themselves near cheap sources of electricity will have a cost advantage, which again increases the returns to centralized manufacturing.

I guess my objection is that when I was younger, we owned an actual printer, and now, we don't. They're annoying. I think the more likely outcome is that you take your design to the "print shop" and get it printed. 3D printers likely require a lot of maintenance and calibration.

In my field (medical physics), the technology is constantly improving, but the maintenance requirements never go away. High precision requires high effort; high complexity, generally, requires high precision. That goes triple if you want to eat off it.

Plus, you probably want a variety of materials — are we going to eat off of a plastic spoon, or melt metal in our houses? Space Kinko's can stock everything from aluminum bronze to Zylon composites.

The copyshop model might work out really well. Perhaps cross-pollinated with the MacDonald's model, ideally not by making the entire site run under e.g. Shapeways flag but less granular: "Mike's materializers" around the corner might have brand A processes x, y and z available, and brand B processes w and x, whereas "Jen's stuffmakers" further north has u through y from B but only y and z from A. The model directories list compatible sites near you with each model.

Right, we aren't there yet, but that doesn't mean we can't get there.

Growing up we had an Adam computer in the 80's but we got rid of it and didn't have anything until years later when we purchased an Apple LC II. That's what we need–an Apple computer for 3D printing/CNC/Laser. The Shaper Origin is a great step forward, but it still requires specific skills, but I do think we will get there.

Sure, every home could have its own craft beer brewing kit.

The thing is, unless you're enjoying it as a hobby, it's better to just buy it pre-made. You don't need to buy materials to have on hand, or guide the process of making it. And, you don't have to wait for it to be finished.

Same goes for any number of other hobbies- silk screen printing t shirts, candle and soap making, etc.

Would I want enough printable materials on hand to print a couch at any given time, or would I want to order them, then manually print a couch? Or just order one from a local print shop and have it delivered already made?

99% of people don't want to lose valuable storage space in their home to raw materials if they don't have to.

Unpopular opinion: I’m disappointed we put so much effort into designing forks, cups, and spoons.

Are any of these things better in 2022 than they were in 1922 or in 1722?

Let’s pick a standard design for forks and only update it when we get new classes of materials or new manufacturing processes that require or enable a design tweak.

We complain about the amount of human ingenuity that gets sunk into ad click rates or tricking people with dark patterns.

What about the amount of human ingenuity that goes into redesigning a four-legged wooden kitchen chair that looks like a four-legged wooden kitchen chair, or a stainless steel fork that looks like a stainless steel fork.

Design in furniture and certain kinds of cutlery is better thought of as art, not industrial design - they are closely related to architecture in this way. Sure, the piece has to meet some basic engineering needs, but otherwise it's main focus is decorative, not functional.

You may not care how you kitchen chair looks, but I assure you the vast majority people do care, at least as much as they care how their T-shirt or pants look.

This is true for almost anything, for example bicycle tail lights:

https://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Da...

There is an immense variety in the 10-20$ range, with very little brand recognition or true transparency on quality or attributes. It is likely that new iterations repeat previous mistakes or regress.

What we need is to find a robust, simple, sustainable, long-lasting, repairable optimum that is truly environmental friendly and is produced ethically, and then, as you suggest, focus human ingenuity on something else. This will not happen in the current economic system, but will require some kind of intervention or crisis.

By the way, if you're looking for the perfect fork, may I recommend the MSR Alpine tool fork? I bought it out of curiosity about reading an article about how it was designed. Was skeptical at first, but it does have great mouth-feel as well as a couple of other neat quality-of-life things. I use it for daily (non-camping) use btw.

Actually...

My wife just found an amazing set in Thailand that not only look unique and cool, but in fact have important new features. The dinner knives are sharp enough to cut steak. The handles weigh enough that you can satisfied suspend the ends off the table, which means they handle brilliantly.

The list goes on...

Agreed! I'm also bullish on more localized manufacturing in which we have more control over the function, aesthetic and shape of the end products.

I wonder about this often. While I agree that additive manufacturing is incredibly exciting and just getting started, I don't expect that it will get as much mass market adoption as you think. Of my last dozen purchases or so, only a few would have been possible to 3d print. Many include microelectronics or precision machined surfaces with a finish that 3d printers would be hard pressed to match. Several need greasing, and only one was made entirely out of the same material. Nobody is printing entire washing machines or motorcycles anytime soon on house-level printers.

Music, movies and software had the huge advantage that basically all the effort is in the up-front design and then it can be digitally copied at basically zero cost. A car design or washing machine design on the other hand is only a small part of the total effort required to fabricate it. They require at least a dozen different raw material types, careful assembly, electrical certification, programming the microprocessors, greasing the bearings, etc. Most people will have neither the inclination or the skills to do the required post-processing themselves. Anyone living in an apartment will probably also simply lack the space for big machinery, especially if it sits idle most of the time.

Even apart from whether it could be made at all in a consumer-grade printer, some things are just unbeatably cheap with modern mass manufacturing methods. Your example of a cutlery set is one: a modern hydraulic press will stamp hundreds of spoons per minute out of steel plate. That process probably won't improve a lot by transporting the raw steel to your house first and manufacturing it yourself.

I think there are massive opportunities for additive manufacturing in industry, where companies would be willing to spend several million on a production grade device and can hire dedicated operators to get the most value out of it. You can already see that happening in the aerospace industry, and it will probably trickle down to almost anything that requires complex shapes in their assembly process. I don't think it will ever move beyond hobbyist in the home scene, the machinery is too expensive, too big and too complex. That said, the type of person who in the 80s would have gotten a lathe for their home workshop could now get a 3d printer instead (or both!).

I’ve been dreaming of this for years. You can already see these economies starting to open up, like the model market for 3D printing.

Great summary tweet thread here: https://twitter.com/tsungxu/status/1565322844418052098

Thanks for sharing the thread!

Mycelium insulation batts have been a "hope this is available soon" technology for green home builders for at least 8 years (Chris Magwood's 2014 book had hopes for it). Seems like insulation would be the most low tech and first use for it. Hell, we can't even get wood fiber insulation boards in the USA from more than one company at a time.

I certainly materials change, but will I be alive to see it? (til then I'm living like an hippy/hobbit, using plants the old fashion ways).

I have been looking for "what do I do now software seems to no longer be able to chnage the world"

I might be pessimistic today. But this seems like a new eye opener

Robotics is a lot of fun! You can pivot from pure software to a software/hardware job. Lots of process automation, or hardware test engineer jobs at robotics and manufacturing companies require basic familiarity with hardware stuff and a focus on software. This can be a good way to pivot away from pure software. A good way to get started is to get a basic arduino kit, blink some LEDs and move some servos, then get a raspberry pi and make a little robot (without the arduino, for variety). Get a 3D printer and do a little CAD.

All these skills will become useful once it is time to build a motor test stand for a robotics startup's assembly production line. After every assembly step they need to run a hardware test cycle that checks all the sensors. Pure software engineers don't always know how to handle hardware, but with a little practice you can build up those skills.

Once you've taken on one role like this professionally, you can pivot even further to hardware as desired. Good luck!

but then you're switching to less paid career path, don't ya?

This probably depends a lot on where you are in your software career path, but in Silicon Valley robotics skills are in high demand. Most software engineers have trouble crossing the gap in to any kind of hardware interaction, so you ought to be able to charge a premium for this skill set.

I personally am more concerned with doing work I care about, so I've never been too focused on getting paid the maximum, so I don't have the most expertise here. But I worked at Google X Robotics as a hardware test engineer. I was a contractor so I don't really know what the pay packages were like, but I would say that even the people who were primarily software engineers on that team pretty much all had hardware skills too. When you're working in robotics, you tend to at least play with all the stuff I mentioned above, and I would expect this experience to factor in to hiring decisions. And the people working at Google writing software for robotics seemed like they were all highly paid engineers, though I never asked anyone specifics.

But if someone feels like software is a dead end job for them, moving in to robotics is a great way to make things more interesting. At one of my jobs I wrote C++ to help a researcher implement some of his ideas around locating specific individuals in a room, and I basically got to play hide and go seek with a Toyota HSR robot all day long. Working with robots is way more fun and interesting than writing web forms all day.

Software keeps being able to change the world, you just have to look for it :) Many changes have been for the worse, but there are changes none-the-less.

Agreed! It is and will play a big role in current and future innovations in the world of atoms.

if you like software and want to work with companies building biomaterials, check out latch.bio and shoot me an email if you are good at programming - kenny@latch.bio

Thanks, that's how I largely felt too. Software is a key part of developing biomaterials too, but deep tech is so much more.

Dall-E et al are going to radically change the world.

As someone who invested in Amyris (heavily mentioned in these articles) last year with much the same thinking as the author… be very careful about betting against incumbent tech.

> be very careful about betting against incumbent tech

Can you elaborate on why you think 'incumbent tech' has anything to do with Amyris stock price?

Here's what I see just briefly looking at some numbers and slide decks:

Amyris has OK revenue growth and gross margin, but wow do they burn a lot of cash. They took on massive convertible debt back in Nov 2021, are continuing to do M&A, and in most recent earnings report are blaming horrible EBIDTA performance on 'higher freight and logistics and increased SG&A, mostly marketing investments'...

  1. Revenue growth
  2. Product innovation
  3. Management

I think a lot of investors make investment decisions based on #2. The product is great. Next-gen technology. Perhaps #1 follows... But even if you get #1 and #2 right, #3 can make or break the valuation of a company. It's not clear that Amyris has a path to profitability and the revenue growth is far from earth-shattering.

Check out "Mistletoe viscin: a hygro- and mechano-responsive cellulose-based adhesive for diverse material applications" https://academic.oup.com/pnasnexus/article/1/1/pgac026/65494...

You can make a kind of plastic out of milk: https://plastics.syr.edu/page.php?id=/materials/casein It's biodegradable.

Thanks, I'll take a look!

The sci-fi book I'm in the middle of reading regularly mentions a popular residential building technique of planting bioengineered seeds that grow into a permanent structure over the course of a year or so. The chapter I just read described some "homesteaders" training and weaving the growth of the plant over hemispherical balloons.

I first heard of seeds growing into structures from Drew Endy at Stanford. Fascinating concept. May take several more generations of biomaterials to get there.

I've been thinking of this topic for so long but never found where to look. Any other venues to read about this ?

This part two of a series. Be sure to read https://www.tsungxu.com/materials-paradigm/ as well.

Thanks for sharing!

Not any blogs that I know of specifically on performance biomaterials.

But this podcast with 6 episodes is great: https://open.spotify.com/show/5ldzwDpvUsaaQVVjMLRlLA?si=33fb...

Also SynBioBeta has some great content as well: https://synbiobeta.com/read/

Thanks a lot.

Very welcome.

Awesome vision by Tsung here. Excited about the possibilities with new biomaterials

Thank you!

Biomanufactured materials have been here for decades. Beer is a biomanufactured material. Quorn is vat grown mycelium. Some drugs are made from modified yeast. Sauerkraut, kefir, yogurt, kimchi. Lots of foods come from animals and plants, often hybridized or genetically modified. Maybe biomanufactured is not the right word to use here. Something like "synthetic biology" would be closer to what the article seems to be about. Ah, they do mention sythetic biology. Funny there is no mention of quorn though.

This was all very new to me and quite awesome. Are there any good companies to invest in who are going to be riding the wave of this tech?

Thanks! Not financial advice but some of the public companies that offer services in the biomanufacturing tech stack. Performance biomaterials companies are not public yet and may take quite a few years to get there.

>This is a classic example of high value, low volume which often serve as beachheads for nascent technologies.

There's a link for "often" that's broken. It's set to

    https://www.tsungxu.com/performance-biomaterials/biomanufactured%20chemicals/

but leads to a 404.

Thanks, fixed.

I used to do work with chitosan, synthetic nacre, and keratin-based biomaterials. While they will have their direct uses we will need a cultural shift if they are to substantially replace petroleum-based materials.

Nice. I've heard a little about chitosan and am aware of keratin but not as a biomaterial. The case the article makes is that functional biomaterials with superior properties have not been made at any scale but that it is on the way.

The future is now, things keep on improving, and it isn't even hardware.

Agreed. Still very early in the adoption of these materials, and biomaterials in particular

The graphene too, when it was hard to make the handful of experiments on it were already showing odd electrochemical properties. Now that it is less expensive to obtain, I suspect we'll start bumping into all sorts of strange things. One particularly interesting one was apparent high temp superconduction in the presence of certain classes of hydrocarbons.

I've not seen that research or finding. It seems we're super early in the potential of 2D materials, especially as you say, as they come down the cost curve.

If you enjoy this read, you should also read the author’s earlier masterpiece of a post on the energy and materials transition (~1h40 read) : https://www.tsungxu.com/clean-energy-transition-guide/

a quick skim suggests this guide is largely motivated reasoning rather than being objective, as it immediately jumps to solar, wind, and batteries as its conclusive remedy. that's unrealistically simplistic, despite its length. something as large and complex as our global energy supply will need us pursuing every option simultaneously. and the most immediate thing to tackle is coal (being the most polluting, estimated to kill millions per year), for which nuclear needs to be a significant component (being baseload).

Wind, solar and batteries dominating electricity generation isn't a 'jump', why would you think that? That's been the consensus for about a decade.

Only the relative dominance has changed, with 80% being commonly accepted for a few years and 100% now broadly accepted as reasonable.

There are many different consensuses. In this case the most important consensus is what is the opinion shared by the operators of electricity grids and the like, because ultimately they will need to match energy supply and demand.

Their consensus is that specifically batteries are completely nonviable for long-term balancing of intermittent energy sources. Physics simply do not add up.

Batteries aren't used for long term balancing, so that bit is correct. They are great at short term balancing though.

So they'll still use mostly wind/solar/batteries. This is what grids are rolling out right now around the world.

> “And on top of that, electricity production is only ~40% of carbon emissions.”

We are electrifying heating and transport. As a result about 80% of end-use energy will end up being electricity.

That's my point: electrifying heating and transportation will increase electricity use beyond the current 2.5Tw and make it even harder to provision the same duration of grid storage. I added the content in parentheses in case this was unclear.

Higher Electricity demand isn't really a problem though.

The motto "electrify everything" is used. It is not a big surprise that this, and developing nations growth, requires electricity.

That's why everyone is really pleased that renewables and batteries are the cheapest way to deploy electricity generation in history.

Higher electricity demand is a problem is you're trying to. Build grid storage. And no, not everyone is onboard with the idea that battery cost is going to plunge by orders of magnitude. In fact, the opposite trend is happening. Battery costs have increased recently: https://www.utilitydive.com/news/new-york-battery-storage-co...

Your cite says:

> the state's higher storage target and DPS and grid operator support will slash costs to $150-$200 kWh by the end of the decade, based on BloombergNEF estimates.

They have dropped an order of magnitude since 2008, so I'd maybe expect a few more years before it does it again but that does appear to be the current prediction.

Announcing a target is easy, hitting is hard. Especially when the costs of input materials to build batteries increase 4x in a year: https://tradingeconomics.com/commodity/lithium

If their predictions are that costs are going to decrease, but when push comes to shove costs increase above expectations then what does that say about the value of these predictions?

I think he is over optimistic about future and not taking human factor into consideration (human nature, geopolitics, wars, upcoming big crisis around the world because of climate change effects, huge human migrations etc). What he describes is the best case scenario.

right, it's the most optimistic possible scenario, if all of the assumptions (of which there are a lot) are correct and as you point out, all omissions (of which there are a lot, despite the length) are negligible. there are some citations, but the narrative project a certain future way beyond what the meager research suggests.

the article does provide a nice survey of clean tech, but the conclusions should be disregarded.

Nuclear is very important. It's just not scaling fast enough.

I don't see a future where next-gen SMR nor fusion gets to cost parity with renewables quickly or easily. They will have to scale up via beachhead markets adjacent to existing electricity demand sources.

Long term, I do think economically viable fusion will supplant renewables, but that's decades away.

nuclear is not scaling fast enough because it's been subject to 40 years of negative mediopolitical narrative reinforced by poor market and technical regulation. nuclear easily gets to cost parity with renewables when you consider the needed storage for baseload usage and the more advanced grid control variable generation requires.

in the US, had we continued to build nuclear at the rate we were between the 70s and 90s, we'd be at over 50% nuclear for electricity generation, which would have knocked coal completely out of the equation, leaving only nuclear (baseload), gas (variable demand), and renewables (opportunistic generation). over 70 years, fission-based nuclear has caused 99+% fewer human deaths than fossil fuels have.

and yes, there's no need to pin any hopes on fusion right now, which is decades away at best.

If we wanted to build nuclear out like it was the 70's we needed to ensure that 3MI didn't happen in '79. This was the death of the industry in the US. Chernobyl buried it in a lead coffin 6 feet deeper, and Fukushima topped it with concrete. On top of that in the war in Ukraine with fighting around the nuclear power plants has made us that much more aware that the stable political environments that nuclear requires cannot be guaranteed.

We're now seeing nuclear plants in a conventional warfare warzone for the first time in history.

Creators of a Texas plant thought it would never freeze (or that if it did it wouldn't matter with the government's gift of a an extremely small liability cap on nuclear), so they didn't put enough safety stuff for that scenario and had to shut down a reactor unplanned.

What’s baseload? If you mean minimum demand from the grid, in Western Australia where I am it’s around 5% of peak demand, In South Australia it’s zero.

In Western Australia coal will be gone by the end of the decade. In South Australia it’s gone already and generation from gas is on a strong downward trend too.

Australia’s conservative electricity system planning is expecting the country to hit 80% renewables. That’s going to end up being the lower bound.

> What’s baseload? If you mean minimum demand from the grid, in Western Australia where I am it’s around 5% of peak demand, In South Australia it’s zero.

This is an extreme exception to the norm. Usually minimum demand is around 70-80% of the peak demand: https://www.eia.gov/todayinenergy/detail.php?id=42915

Thanks for sharing!