Modeling and the Cloud


By day, I spend my time writing about various aspects of the cloud. And as something of an aficionado of tortured metaphors, for the past month or so I’ve been drawing various connections between what’s going on in modeling and the way that cloud is reshaping how businesses and people do the things they do.

WTF is cloud?

It’s possible to go pretty far down the rabbit hole here, but I’ll keep it as high level as possible. In the simplest possible terms, the cloud is just…computers that are somewhere else. That’s it.

For years, companies (and governments and…) ran their own private data centers. Many still do. These data centers – massive facilities packed full of rack after rack of servers, hard disks, networking gear, miles of cabling, and so on – handled everything from processing financial transactions to storing photos and other media, to hosting websites, to housing customer databases. You get the idea.

Data centers are also expensive. And data centers are finite. Remember the days when a new Game of Thrones would air and HBO Go or HBO Now or whatever they called it then would crash? That’s because HBO’s data center couldn’t scale to meet the demand.

The cloud basically takes that entire data center idea and outsources it. Instead of having to build and maintain and expand its own data centers, HBO now hosts it shit through Amazon Web Services (AWS). So does Netflix. Many of the apps and services you use every day run through the cloud – probably either AWS, Microsoft Azure, or Google Cloud Platform.

Now…cloud is exactly like data centers, except that it’s completely different.

First, the cloud is networked. You’re not talking one data center. You’re talking a network of data centers, spread across multiple regions and availability zones, connected by what’s basically a fearsome, crazy fast private super internet. Need to house data locally in the EU because of GDPR regulations? With cloud, it’s a click of a button (well…maybe a bit more than that, but you don’t have to build a whole damn data center). Everyone binge watching some show about a crazy redneck and his tigers? Seed the video file across multiple regions.

Second, the cloud has a bunch of services built in that let companies take advantage of its unique characteristics. Unlike a data center, compute and storage capability on the cloud is for all intents and purposes infinite (if you’re willing to pay for it). This means you can scale up immediately so your shit doesn’t break under massive demand, for example. Increasingly, these services are being bundled and automated. Terms like Platform as a Service (PaaS), Infrastructure as a Service (IaaS), and Infrastructure as Code (IaC) get thrown around, as does the misnomer of serverless computing. Serverless isn’t serverless at all. It just automates provisioning, scaling, and so on.

Third, with cloud you only pay for what you use. To go back to HBO again, that means they only have to pay for all servers and storage and so on at a massive scale precisely when a new show is released. No need to massively expand a private data center to handle a massive cliff in demand for a few days a year.

Undifferentiated tasks

One term that gets thrown around with cloud is undifferentiated tasks. This is a fancy way of saying “tedious work that doesn’t set you apart”. Netflix isn’t a data center logistics company. Neither is Capital One or Adobe or…you get the idea. Customers don’t give a shit about data centers. They give a shit when they can’t pay for something or can’t watch something. The cloud offloads a bunch of undifferentiated tasks.

In lazy personal terms, this would be like if you could pay someone else to do your dishes and laundry, get gas, organize your mail, do the grocery shopping, and so on, so you could focus on the stuff that sets you apart.

That’s exactly the bet that companies make on cloud. There’s the boring financial aspect – hoping that the pay for what you use model ends up saving money over building and maintaining data centers. But there’s also the bet that focusing resources on stuff that sets them apart will make companies more money, and make the whole thing worth it. And hey – when Netflix can expand its service to dozens of countries around the world on the same day and have it go off without a hitch? I’d say that’s worth it.

And the connection to modeling?

So what does any of this have to do with modeling?

Well, the cloud has already had a significant impact on modeling. Posting your stuff online? It’s almost certainly being hosted and stored in the cloud. Reading this? Same thing. Buying stuff from online retailers, or via eBay? Same thing. But that’s all underlying-gears-of-the-interwebs stuff that nobody cares about.

What I really wanted to focus on was that idea of undifferentiated tasks.

Modeling is packed with undifferentiated tasks. Gluing parts. Sanding seams. Cleaning your airbrush.

What do good kits, good aftermarket, good paint, new weathering products and 3D design and printing all have in common? They remove undifferentiated tasks.

Buying a metal barrel for your tank saves you from having to remove mold or join seams from a plastic barrel, and may also save you from fixing warping. A good kit can get you through assembly smoothly, increasing your likelihood of finishing the build (doubt me – see the explosion of F-14s built after Tamiya released their kit, or P-38s more recently). Good paint behaves predictably, reduces frustration, and renders a lot of the silly rituals people go through to make garbage paint work completely pointless.

Like cloud, aftermarket also offers you characteristics above and beyond the usual. Detail most of us would be hard-pressed to replicate even if we had the time and inclination (which most of us don’t). I mean…I just restarted work on the Trumpeter A-6, and just look at the fabric texturing on these True Details seats.

I don’t even know where I would begin to pull off something like that.

This extends all the way down to basic tools and consumables. Tamiya Extra Thin and other fast liquid solvents let you weld plastic together with little mess and little need to clamp stuff together and let it set for days. Good abrasives make sanding more efficient. A good saw enables you to make cleaner cuts, with less bodywork on the back end.

Vinyl cutters let us design and essentially print our own paint masks. In the past, that kind of thing had to be done by hand, with a sharp blade. 3D printers and 3D design software lets those with the patience to learn their mysterious ways design their own parts. And some of the work happening in that space is truly amazing.

The legacy mindset

Legacy thinking seems pretty endemic to the human race in general. It shows up in the old axiom about always preparing to fight the last war. It’s been the death of armies and empires time and time again. In my lifetime, it’s popped up over and over again. Kodak, Blockbuster, Borders. You could say it infects politics for sure – we’re still to a large extent mired in the same boomer fights from the 60s.

If cloud has a main opponent, it’s the legacy data center mindset. Even though there’s significant functional overlap, cloud demands a different vocabulary and a different mindset. It’s a massive simplification, but I tend to think of it as…the data center is all about managing your control over things. And the cloud is more about managing giving up control. The data center is the guy with a home theater and a coffee table strewn with five different remotes. The cloud is going through the one-time exercise to integrate everything, and then just getting to say “Alexa, turn on the TV”.

But giving up that control is uncomfortable! Changing your thinking is hard work. A lot of companies try moving to the cloud, run into problems, and before you know it, the status quo is back.

To use an example we’ve all seen in modeling. Modeler buys airbrush. Gets marginal results. Uses airbrush less and less. Becomes one of those people on the internet who talks about how airbrushing is too much work. The kind that makes those of us who use our airbrushes for just about everything we possibly can scratch our heads in bewilderment.

What this all boils down to

While all of this has been sort of bonking around in my head over the past few weeks, I’ve been struggling to sort out what it all amounts to. What’s the “so what?” of it all?

And I think, maybe, I’ve got a handle on it. And possibly a handle on a lot of the tedious drama that happens around this hobby, and around most human endeavors.

I think it has its foundations in legacy mindsets compared to those more embracing of change. But more than that, I think it comes back to the idea of undifferentiated tasks.

For some, there’s almost a love of those aspects of modeling. Of filling and sanding and every little step. It’s almost an artisanal approach. For others – and I count myself among them – there’s a desire to focus more on the aspects of a build that interest me, that let me put my own stamp on things. My personal area of interest is painting, weathering, and that jazz. I know others who love scratchbuilding, or who are falling down the rabbit hole of 3D design.

When I stop and think about various dramas of the recent past, a lot of them seem to break along this line.

7 Comments Add yours

  1. Jon Bryon says:

    Interesting post and makes sense to me. It’s just that my undifferentiated task is weathering and the bit I really like is seam removal! Once I’ve got the basic paint on I lose interest, want to finish it and move on and build the next model…

  2. Torbjörn Hanö says:

    Well put! It’s a new world, don’t fight it, roll with it. A Chinese proverb I try to live by is if you are afraid to take the next step, you might find yourself standing on one leg for the rest of your life.

  3. Nicholas Pirnia says:

    This was an enjoyably cerebral post and with an interesting connection to your last post about judging at IPMS events. It seems that IPMS uses these undifferentiated tasks as a filter instead of placing the undifferentiated tasks at the center of their assessment. It strikes me as sort of a means to decide “who is most invested” by placing a lot weight on these undifferentiated tasks instead of focusing on the more individualistic aspects of the hobby. On the one hand, I see how this democratizes the competition in that modelers who are dedicated, but not necessarily highly skilled may still end up in a favorable position given these rules. But on the other hand, why create rules that may exclude master modelers by focusing (or using as a threshold) these undifferentiated tasks? It strikes me as an example of how well-meaning intentions are shaped by committees and practical applications to undermine process of awarding greatness in the hobby…. Oh well, at least it happens in lots of other places too. 😛

    Great post!


    1. Nicholas Pirnia says:

      Quick edit: I meant to say that the IPMS should place “differentiated tasks” at the center of their assessment, not “undifferentiated tasks”.


  4. Isn’t this a discussion about wether or not using new tech is actually still “modelling”? Like I saw 3d printed, pre-painted cockpit instrument panels. Literally, there is nothing left to do but snip off the part n glue it on. I don’t care, it was ww2 aircraft, which I don’t do and I fucking hate cockpits. But I guess the philosophical question comes to my mind. At what point does the removal of old , time n energy sucking processes, get to the point where your really not building n modeling anymore? They could just 3d print the whole plane now and just leave it unpainted, I that still modeling? Is not anywhere near that point for me, I just see great strides in tech that make the process mire enjoyable.

  5. Bob says:

    Good thoughts,
    People in general in this hobby are open to new technology to make there lives simpler, there is a money issue on some of them, alot of people can’t afford a 3d printer, or a vinyl cutter, or some even most aftermarket parts.
    the Undifferentiated part would be better explained, i think, is.. I payed 50$ for this Me 262, now there is a resin cockpit, detail set for it for say $20 more.. now we have to think.. is the 20$ worth it? Am I going to see the cockpit through the canopy? is the in box version crap or decent? pro’s and cons based on personal opinion. Is paying $20 to get a better cockpit worth to me more than maybe a extra hour detailing the kit one? Personal opinions of what it is worth to me.
    As for the preprinted, if its the whole thing printed and you do nothing? like an Eaglemoss spaceship, take it out of the box and display it? then no, its not a model, its a pre painted display piece. now taking a pre painted cockpit and put it in a model? thats just up to personal preference..

  6. Igor says:

    Interesting post, very appealing to me as a software developer and cloud enthusiast.
    One other aspect to the cloud is low price of failure. Start small, fail fast, infrastructure in the cloud is cheap – you can easily trial and throw away things. I’m not sure if that’s the case with aftermarket and painting – failure can lead to hours and hours of reverting the damage, or even abandoning a build. The “undifferentiated payload” of aftermarket and weathering product is sometimes cheaper than time of DIY job. But it is a hobby, not building a system to market as soon as possible, so is time that critical?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.