Episode Thirty One: Everything’s Made Of Blocks; How The World Works; Advances In Quantification

by danhon

1.0 Everything’s Made of Blocks

I have a secret confession: I’m not that into Minecraft. I appreciate it, sure, and understand it (I think) and its appeal, and, for the record, *some of my best friends are into Minecraft*, too. But other than having bought it a long long time ago and occasionally wandering onto a friend’s server, I haven’t spent that much time with it. (The last time I remember looking at something because it was a genuinely interesting cultural phenomenon, i.e. for ‘research’ was World of Warcraft and I only managed to escape something like 67 levels later. It may well be that I’ve just developed better self-preservation instincts.)

All that is to say that this is clearly an Olympic-level reckon, in that what’s interesting about Minecraft is its accessibility (not to be confused with *simplicity*) and its depth. Now accessible things that can be deep often appear to be impenetrably complex from the outside – but (and this is why certain games are so interesting) it’s precisely that combination that allows for a long-term and rewarding experience.

Minecraft, a simple combat/building game feels like the real-world incarnation of OOP!, the product of Interiority, the startup team detailed through Daniel’s journal in Douglas Coupland’s Microserfs. When Microserfs came out in 1995 (the year I was excited by a Microsoft operating system, and the year I turned sixteen), I devoured that book like I devoured that other Douglas’s book, The Hitch-Hiker’s Guide To The Galaxy. Based on a short story Coupland wrote for Wired in 1994, Microserfs inevitably became the textbook for this introverted geeky teenager – and it didn’t help that I shared the same name of its protagonist – almost a how-to manual for how I wanted my life to turn out. But I digress.

Minecraft is the building game that LEGO would hope to be if it were incarnated digitally. Blocks, yes, but oh-so-many. And combined with a crafting system, I can see how if I were an eight year old right now, I’m be totally into that.

Anyway: you know that you’ve got a good thing going when people are subverting your original platform and turning it into things that it’s either unsuited for, or just plain orthogonal. This isn’t completely off the walls, *but* the thing that caught my eye was the story that PhD student Christopher Mitchell has a project called Sparseworld[1] which is using off-the-shelf (ish) open datasets from the US Geological Service, Google and OpenStreetmap to programatically recreate Manhattan (and, of course, once you’ve programatically recreated Manhattan, the world is literally your Minecraft block-recreated oyster). This is *fantastic* because you can a) see it right now[2], and b) he’s using a server cluster of 300 cores and 200GB RAM to render Manhattan which is pretty achievable for anyone with a credit card and an Amazon account these days. But it’s being done in *Minecraft* – not Unreal or Unity or anything. *Minecraft*. Which was never really intended, I think, to do these sorts of things. But is a wonderful example of the perversion of a just-good-enough and accessible tool.

[1] http://www.cemetech.net/projects/item.php?id=47
[2] http://nyc.cemetech.net:8123

2.0 How The World Works

This might sound a bit like wankery, but I think we’ve reached the point in society now where our reliance upon networks and computerisation have pretty much forced upon us a new layer of understanding how the world works. This isn’t the “Learn To Code” model of how the world works because it feels like learning how to code, at least in the English model of “here’s how we make a website” or teaching Jquery is nothing more than an updated version of “learn how to use Microsoft Office” – because you don’t really need to learn *how* to code, but need to learn *what* code is, and what it does.

So here’s (possibly) an ongoing series of concepts that I reckon are required to make sense of and understand how the world acts the way it does, especially when that world is touched, repeatedly and often without consent, by computers.

Computers do what people tell them to. They are exceedingly dumb. This means that we can make them do exceedingly dumb things, or exceedingly smart things, but, and only but: they only really act in or towards what we tell them to do. Now, this holds true because for now, we don’t have self-modifying goal-oriented programs (right?) – because even when Google accidentally builds a neural visual system that identifies cats, the people involved were still *intending* to build a system that recognised features in YouTube videos. They may very well not know *how* it recognises cats, but that’s by the by. Similarly, we *tell* computers to execute Flash crashes. They may well be unintended consequences, but that’s not to say that they weren’t *contemplatable* consequences.

So the thing about the internet not being secure and everyone freaking out about what the NSA and all our other nation-state (and corporate) organisations are doing is this: we *told* computers to send information over the wire, in the clear. That was *by design*. The bitstream and the clickstream are clear because a human made a decision somewhere and that propagated out over the network. We have to remind ourselves that we have a remarkable degree of agency, still. There might be different reasons why, for example, we decide to retain everything ever (from corporate data-mining for profit, to an instinctive packrat tendency) but the fact of the matter remains that ultimately, computers do things that we tell them to do.

The corollary to that is being aware of the infrastructure that the network is built upon. It’s so unbelievably complex. You wouldn’t normally (or would you?) want to worry about which particular part of the OSI 7 layer model has been suborned. Or that, as it turns out, one person on one mailing list could have an opinion that turns a special interest group in a certain direction which means a certain implementation of a certain cryptographic standard gets implemented into standard libraries.

Here’s the thing. You can *just* about understand how complicated the Space Shuttle is. You can look at it, especially as it moves from its hanger on that awesome caterpillar thing over to the launchpad and then you look at Mission Control and there’s all those complicated people there and remember that movie Apollo 13? That was some crazy stuff! All those people, working so hard!

The internet is *harder*.

The Shuttle gets props for being, I think, regularly named “the most complicated machine.” With all due respect, bullshit. The “internet” is the most complicated machine. From the physical layer through to the presentation layer that’s some dude using BroApp to send a text message to his unfortunate girlfriend, *that* is the most complicated machine we’ve ever built. Because with the Shuttle, we could fix the fucker. There was that story about how we might, just, could’ve saved Columbia by NASA mounting the most daring rescue mission ever. If we found a similar flaw in the internet? Goodbye internet. We can’t just do a code audit of every machine connected to the network. We can’t just turn it all off and on again. The entire thing is just like one massive, distributed, single point of complicated failure.

I guess what that point is getting at is this: the internet is complicated. Although the answer to some questions about coding are, in some respects “easy”, the implementation is nigh-on difficult. So when someone asks if you could make a triangular web browser, the answer is “well…” because, *in theory* the thing about code and computers is that you can more or less literally do whatever you want. It just depends on whether you want to build an entire toaster from scratch or not.

We haven’t had, I don’t think, an infrastructure failure of (not on) the internet yet. I don’t know if that’s because of its slimy tentacle-y design, or just because we’ve been extraordinarily lucky. But it feels incredibly brittle to me.

3.0 Advances In Quantification

Friends-of-this-mailing-list Product Club[1] released the first Product as a result of their Club today, Up Coffee, by Jawbone[2].

A few observations.

One: someone has finally released something that makes use of quantification that actually feels like it’s *useful* or solves some sort of user need (I want to sleep better) other than “I want to measure things because I’m into measuring things” or “I want to lose weight”. Or, at least, what I mean to say is this: Up Coffee has a point to it that is something other than intrinsically just measuring something to see what that says about the something you’re measuring. It is about coffee, and caffeine and sleep. Thank God for that.

Two: it makes me yearn for a *passive* personal trainer model. Nike’s new FuelBand software has this thing you can turn on which reminds you Win The Hour, which is great if you’re trying to remember to be active throughout the day, but if you’re like me and British, there are only so many times I can be reminded to crush it before I have to crush something else.

The thing about a lot of these bands and quantification devices is this: they’re great at calibrating behaviour and setting in motion a new habit. They worked great at that for me. But then, when you have the habit in motion, you kind of don’t need the persistent monitoring. It’s almost as if you need some sort of “OK, you’re in the rhythm of it now, I’ll only disturb you *when something drastically changes*.”

Almost as if, after a while, the default view would be wide and would pretty much say: “You know what? You’ve been OK for the last three weeks. Nothing to see here! Keep it up!” because really – *there’s no need* to see all that data. It’s just stress inducing. God, imagine if your legs emailed you every day to tell you how much you’d used them. For your entire life. Shut up, legs. I’ll make a rule for you and not listen to you.

This is the thing about personal quantification software that is like a shrill little demon sitting on your shoulder and never shutting the fuck up. You end up hating and resenting it. A good friend, a good trainer, knows when to back off and when to be relevant. What we have now is very far from that. And really: what’s the use of all this big data if you’re just going to *show all of it to me*. Seriously.

[1] http://productclubsf.com – still suffering from “everything else is more important than our own website” syndrome
[2] https://jawbone.com/up/coffee

Instead of a sick baby, I have a baby with a sick mum, so we took the baby to the play gym so the sick mum could rest. In my mind, the play gym was a wondrous world full of Things, but in the reality of Parks and Rec, it’s a gym that babies can play in. My baby’s awesome, though, so nyah.

Thanks for reading, and as ever, please do send me notes, even if they’re really short. Or even if they’re really long. I reply to both kinds. Honestly, I do: people who’ve received replies have been surprised. I like surprising people.