DS062: A brass fox’s skull sits on a desert floor with a sandstorm blowing around it.
This week I was alarmed to discover that Gifs are for Boomers and generally frowned upon by gen-z for being lazy in a time of hyper-personalisation and custom content. Some sort of threshold has now been passed. Anyway, another short one this week.
DS062
Last week’s sketch was debilitatingly difficult and frustrating so it was hard to be motivated this week to do anything too complex or mechanical. As a consequence, this is a super simple sketch with nothing particularly complex going on. Procedural sand for the ground, downloaded fox skull rendered in a brass material then a volume with a noise texture animated moving across it to simulate wind. I suppose thematically it’s more about being lazy and quite exhausted by last week’s sketch.
The size of last week’s sketch also means I lost the video in a cloud syncing crisis but I actually don’t care. I have an idea for some new abstract style to try out this week. Quite looking forward to that.
Recents
I’ve done a couple of podcasts appearances recently that should be coming out soon. Obviously I’ll share when that happens. On Monday next week I’m doing a talk for the MFA Transdisciplinary Design folks at Parsons which should be fun. Something around imagination and expectation and hype I think.
Short Stuff
Really good and thoughtful episode of Interdependence with Kyle McDonald with a nuanced breakdown of Ethereum energy costs. I think they misattribute the backlash on crypto to being about professional rivalry but it’s a much more balanced perspective. McDonald puts really well some of the issues with replacing publics and public funding with crypto and mistaking hypercapitalism for egalitarianism .
You know what, best inventions of 2021 from Time. Not a bad list at all. Warning: There’s a lot.
‘Meta’ aka Facebook, have filed a bunch of patents for face tracking and sentiment analysis for the ‘metaverse.’ In case you were under any illusion that the genuine intent was to foster closer community ties, it seems mostly related to increasing the nuance and volume of targeted advertising. Which ever version of Web 3 we get; whether crypto or meta, it will be profoundly disappointing.
Super brief I know but at least it’s here. Love you.
DS061: A dimly lit room with three large frames in. Each frame contains a suspended weighted pendulum with a sandcastle bucket on one end. The bucket fills with sand from a pipe at the top of the frame then falls to the bottom where it ‘drops’ a sandcastle on a tray table at the bottom of the frame. The sand is pumped out of the tray table and back to the pipe at the top. The room also contains a folding chair and table and file boxes.
I’ve been really struggling to not read stuff about Web3/Crypto; it’s like rubber-necking a car crash. I can’t tell if it’s just that there’s a lot of it about or if the algorithms are pushing it my way. I’m going to try and tone it down and share less on it because I sometimes worry that there’s some perverse incentive at play. The more brilliant, thoughtful, critical and analytical pieces being written and shared about how it’s an extractive pyramid scheme to enrich a super elite the more attention is driven to it.
I’m also going to apologise because there’s not any ‘content’ this week. Just a lengthy breakdown of the sketch and then short stuff. To be briefly hypocritical and write about it, I’m not the only one worried about this notion. As the latest Real Life Mag newsletter attests:
Whenever I write anything about “Web3,” I am beset with the self-defeating feeling that the concept is impervious to criticism. Not because it is any way defensible — its success will make the world immeasurably worse in every conceivable way, more divided, more unequal, more paranoid, and more unsustainable — but because the concept seems likely to profit from any attention it’s given.
This was a nightmare – please skip past this bit if you’re not particularly interested in me moaning about my own hubris. I started this week’s sketch with the thing I’ve started doing more often – building a set and then figuring what should go in it. I had the notion of having a neoclassical room that would be lit like the interior scenes in Dune. Turns out that’s pretty tricky but we can skip over the details to the animation.
I was giving a talk about worldbuilding this week so that was running through my head while doing this sketch. I thought about a machine that would endlessly build sandcastles only for them to collapse. The set-up for this simple animated machine is extremely complex: The rocking arm is keyframed, copying the movements of a Shishi-Odoshi video I found. Getting the pouring sand looking right was the next issue. At first I tried particles, then liquid before using a smoke simulation with gravity flipped and multiplied so it went straight down. If you’ve ever played with fluid simulations of any sort you will know that they are an utter pain and getting a very specific look – that of pouring sand, even trickier. Then I wanted (for some reason) to have multiple sandcastle machines so I needed to re-calculate fluid simulations for another two ‘stations’ resulting in about 2 GB of simulation data and maybe 48 hours of computer time?
Then I needed the collapsing castles to line up with the arm. I ran through a bunch of methods here for hours; particle dissolving, fracturing, fancy plugins that turned it into molecules etc. etc. All of them ran the computer into the ground when totted up with the fluid simulation. In the end I decided to be pragmatic after realising that on-screen the castle is relatively small. So it’s simply a deformed mesh that it flips between ‘fully formed’ and ‘flattened.’ Then the castle mesh jumps back into the bucket and falls down again with it. I will say this: every frame of that castle moving with the bucket is hand animated.
So that was about three or four days work, but then came the real problem; rendering. I built it for Cycles knowing I would need Cycles for the flowing sand – Eevee can’t really do volumes or translucency well, both things that are central to this sketch – but Cycles was taking approximately 45 minutes per frame to render because of the sheer amount of stuff going on – so about 5 days in total to render. No good. So I had to switch out everything for Eevee which meant redoing a bunch of shaders, lighting etc. only to find that the amount of data being crunched resulted it in dropping caches or randomly crashing. So I baked everything and rendered it in 50 frame chunks by hand over the weekend.
Lessons learned? Don’t fly too close to the sun. This was too ambitious to do for what’s meant to be a sketch and if I’m honest I don’t even like the way it looks that much. The aim of these sketches was to do something that maybe takes an hour or two a week and this one took damn near every free hour I had.
Working through the difficulties of this week’s sketch. L-R: Initial drawing to figure out how all the parts of the fluid simulation, the dissolving sandcastles and the key-framed animation would fit together: Eventually settling on downward smoke as the best way to simulate the physical properties of flowing sand: Working out how the castle mesh would dissolve using shape keys.
Recents
I was up at the Architecture Association last week to talk with some of the students about worldbuilding. The talk focussed on the ways in which worlds are conceptualised through the tools use to measure them and then how they are reproduced, with a particular focus on (and you’ll never guess) AI and Computer Graphics. The slides are up here if you want to see.
Short Stuff
Brad Troemel’s NFT Report is online for free. A forty-minute breakdown of what happened in art, tech and finance to create the Ponzi scheme of NFTs. It’s pretty good and concise and in Troemel style has some nice gags.
The Expanse ended last week. Easily my favourite TV show of the last few years, and likely yours as well. Everyone knew the ending was going to be rushed due to the practical constraints of TV production but the vibe generally seemed to be that it was ended tastefully. Post-mortem writing is beginning to appear which is great. Here’s a piece from Noah Smith making a case for it as an optimistic (or journey to an optimistic) future.
Mrs Revell said to me that she likes it when I used to finish on personal notes and I don’t do that so much anymore. So what has been going on? Well, we had a daughter in early December and I’m currently on a sort of three month-ish sabbatical with her and Mrs Revell, doing deep ethnography on poop, hiccups and flailing arms. Both ladies are very healthy and we’ve been incredibly lucky with how well everything has gone and how chill Mrs Revell Jr. is. So yes, I’ve been generally being domestic, reading and writing in the morning when I take the feeding shifts while Mrs Revell sleeps, watching a lot of Queer Eye and riding bikes. It’s great because I haven’t really had a proper break in ten years.
At the end of February I’m starting with Arup Foresight and I’ll start getting warmed up for that at the beginning of the month.
Ok, love you very much, but you know that of course. I still like telling you, I think it’s nice to hear even if you already know it in your heart. Speak to you next week.
DS060: A petrol (or gas) station at night. Two cars are hung from a steel ibeam and rotate slowly around the middle of the forecourt, like a big car mobile.
I finally got around to setting up the subscription service to this blog. So if you’d rather get these posts by email you can sign up down there ↙. Unfortunately, the animations won’t embed but you’ll get a nice still image. If you’re one of the fifty-odd folks already subscribed then welcome! Last week was a bit of a test, please send me any feedback on the email version. You can click through to see the animation, or they’re all on Instagram.
DS060
This week I decided to play around with some methods for photorealism. To be clear, the cars are found models from CGTrader already textured but the rest of the scene is done by me. Cars are whole projects in themselves but useful assets and I’m not into them enough to start modelling them by hand. I used various photos from petrol station from the internet to do some project-from-view UV mapping of the pumps, bins etc. similar to this method. There’s a big cube volume in there for depth and mood and then straightforward lighting. Following last week, I’ve found that adding in noise post-render undoes some of the painterly-ness of the de-noisers. All in all actually quite straightforward this week.
Prognostications
I missed this piece way back in 2020 – the promissory year of the self-driving car – untangling the complexities and hype to explain why we still didn’t have self-driving cars (spolier: we still don’t.) It’s rare to get excited about the future now, I think, for most. There are glimpses of it around elections and announcements but it often disappoints. I can never tell if it’s me that’s changed or the world, but the hysteria around a new iPhone or service seems to have vanished a little from the world. The excited word-of-mouth chatter for a new app or gadget has been replaced by rabid zealotry met with bitter dismissal. I’m looking forward to the remastered release of Witcher 3, though.
Anyway, rather than do a year-in-review of 2021 (which sounds like a lot of work) I thought I’d trawl some things from some of the articles that proliferate around this time of year about predictions for the coming twelve months, adjust or add my thoughts.
Fusion and Quantum enter the rhetorical arena. Azeem Azhar highlights the dramatic increases in funding for both quantum computing and nuclear fusion that 20/21 have seen. He did a great podcast episodes about both these technologies here and here. These are both technologies that seem reasonably exciting in actual application. Based on the financing alone, these seem like reasonably sure predictions but I’d add that I would also expect them both to get sucked into the rhetoric-hype vortex of silicon valley as a result of the increased funding interest. The tragedy is going to be seeing them put to the service of social networks and/or so-called AI in the pursuit of power. We might expect to see ‘quantum’ in particular dropped more and more into advertising and pitches to boost hype while (like AI) the technology itself is still in a very formative state.
You built it and now they’ve come: Crypto-exclusionism. Quite possibly the emergence of a social division between crypto people and non-crypto people. My views on this are pretty-well known if you read this: It all seems to be a technology that, despite the feverish defences put up, is being adopted by those in the relentless pursuit of the accumulation of personal wealth. The grotesque and rapid accumulation of personal wealth is regrettably one of the greatest drivers of the adoption of new technology and so crypto is now probably here to stay. (About ten years ago I mined SolarCoin in our studio and that was my first and last dabble.) A good test is to ask someone to explain crypto to you without mentioning money. I’m always up for hearing another argument or rationalisation but I have yet to hear it. Either way, similar to early social networks, as services become crypto-normalised, those who don’t engage may find certain services and products out of reach. Tech-exclusionism has impacted transportation already pretty severely in places where Uber et al have taken root and so the spread of crypto across digital transactions will probably exclude non-adopters; non-adopters of choice (like me) and those who are on the other side of the digital divide. Cryypto, like ‘digital’, ‘social’, ‘connected’ will just be in things without the need for its own prefix. Crypto will probably continue to be incredibly divisive in a way that things like smart phones, self-driving cars and even so-called AI, aren’t. You see, even the most fervent luddite can see the potential application of so-called AI in faster and more accurate medical diagnosis or climate modelling even if we need to scrutinise and critique the methods and tools. But with crypto, the voices of people working on more than the money need to be amplified. Tangentially, an even darker prediction is that the inevitable crypto crash will lead in the deeper entrenchment of right wing extremism, particularly in the US.
No rise of the machines. The Wall Street Journal has put together its predictions for next year. Nothing particularly surprising: More gadgets and gizmos aimed at dealing with manufactured inconveniences than anything meaningful. However, there’s a peculiar leaning on robotics in the domestic sphere, both delivery bots and chore-bots. I really struggle with seeing the story of these machines; why would anyone want them? They can move things around, sure, but they require a bunch of extra work from users to look after, charge, store etc. The dream of robots in the home is almost 150 years old and I don’t think 2022 is the year we’ll see it manifest.
Curiously muted AI expectations. Unusually, this report from the AI world implies more humility and pragmatism in the gradual, incremental developments of the field. That doesn’t mean the hype will go away but perhaps part of the normalisation of AI is it’s sublimation into other processes and the background noise of general technological innovation rather than big blowout AI-specific announcements. As above with crypto, we’re less likely to see ‘smart’ or ‘intelligent’ highlighted as advertising gimmicks as they’ll just be expected to be part of things in the same way ‘digital’ and ‘connected’ are.
Planetary Design. Alice Rawsthorn put it beautifully here. I think it’s increasingly going to be accepted as a mainstream approach to design in practice to think on planetary and giga scales about design interventions in the environment as well as the role that design has in communicating and conceptualising large systemic change. There’s been a flurry of new degrees established in the UK to corner this demand and work I’ve done with clients over the last year has shown that non-designers are developing an understanding of the potential.
Gamification of Fitness. It’s a little like smart fridges; there are gradual, incremental improvements but you really don’t need a gadget to know you ate too many doughnuts any more than you need a notification to know you’re out of milk. For some, maybe data does change behaviour (I know it works with my cycling) but in this prediction, health data is artificially loaded with financial value which reads a little bit like the Chinese social credit system but for corporations to profit off. The hope is that by being fitter, you become more ‘valuable’ in the most literal cosplay of Foucauldian biopolitics ever instantiated. It’s unclear why you would be more financially valuable by having a fitter or healthier dataset when everything from private health insurance to diet plans to medical research relies on you being unfit or at least paranoid about your fitness to profit.
Lifespan and reusability.This is a bit of a no-brainer. Partly due to ‘reading the room,’ partly because of actual legislation and partly because of the material issues in the next point, we may be seeing the end of the age of planned obsolescence. This has all sorts of implications for insurers and manufacturers but perhaps we’ll see them begin to tackle the enormous wastage of gadgetry. I would expect that this year or next we’d start to see lifespan and reusability as a marketing pitch on mainstream devices. Repair and reuse might also start to enter design curricula not as a novelty but as a fundamental.
It’s the coke, stupid. We had inklings of this with thing like the Evergiven crisis but 2022 could really be the year that the hidden material cost of tech starts to bite. The cost of silicon and chip manufacturing has been increasing through COVID but has also been affected by the environmental cost of turning Bauxite into Silicon crystals which requires huge amounts of coking coal, mostly supplied by China. But in other areas too are similar issues, a European requires about 15 tons of steel – also a coke-intensive process and other countries are catching up with this demand. There aren’t currently alternatives to the manufacturing processes of key materials like silicon and steel so the simple response is to stop using them. Canon has removed DRM chips from printer cartridges, ironically packaging chips with instructions on how to bypass the DRM.
I’m biased of course because my gut tells me that people are ready to turn away from technocentric solutionism and that there’s a popular and growing mistrust of Silicon Valley to deliver the utopias it’s promised as demonstrated through (for one example) the succession of issues that Facebook has faced but also the rising cost of living due to material and energy costs. Anyway, I won’t review these at the end of the year because they’ll be wrong.
The future is not only useless, its expensive. Well-written description of the dispiriting vacuousness of crypto culture. It’s a bit snobby on culture – I firmly believe that people like what they like and that their right – but the general notion of financial exploitation is great.
Again relatedly, Brad Troemel’s NFT Report, online now for free is a really great analysis of the bubble.
This report on the state of Web 3 has been getting a lot of attention. The tldr is: Web 3 is just Web 2 dressed up in clothes with less privacy but people are making money off it so they don’t care.
That was by no means everything. I’m doing an interview in two weeks where I’ve been asked to assemble some ‘weak signals’ – are there any ‘weak’ signals anymore? Feels like everything is pretty loud. Anyway, you know what I’m going to say. – that I love you. It’s my affectation for these blogs but it’s always important to let you know, I think. Also, my time is a little bit freer at the moment so let me know if you’d like to catch up, in person (fully boosted up here) or via teleconferencing. Have a great week.
