BOX038: JUST BECAUSE IT’S BIG DOESN’T MAKE IT MORE REAL

12.05.2112.05.21

DS036: Took an old scene and just decided to run a massive fluid simulation in it at 1500 resolution. Took like four days to run the calculation and just isn’t that good. I think it went out the upper-bounds of good fluid simulation. I think I wrote about this somewhere – you don’t get closer to more accurate fluid by increasing computation, rather, the software is designed to work with a certain poetic distance with relatively low resolution. What’s a good metaphor? Like; holding something really close to your face doesn’t make it more realistic. Putting it in the right context, with good composition and lighting does. Also true for rendering software.

At least the warmth is bleeding in through the crack in the sky.

Well, I finally finished and uploaded that lecture. That thing had been haunting me. I’m not entirely happy with it; it’s not such a linear story to connect AI, imaginaries and social fictions – they sort of mutually feed into each other. I’ve been gnawing over the same thing in PhD work. With both text and video you’re tied into a linear narrative. Without a neat historical trail or clear cause and effect it’s really hard to sort of lead people through your thinking. I also didn’t have time to put in all the gags I wanted to, video production is hard and very hard on your own. Anyway, it’s out there and I’ve done it and now I will keep learning.

It’s been nice seeing people start to mix with the format of talks. I mentioned some last time from Crystal and George but I guess it’s a thing now. Festivals and events have been dying off for a while and I’ve always seen talks as a way to try out talking through ideas so it’s nice to see that that aspect of it is still alive and kicking. Just not again for a while please. Anyway, amidst this I had to pull out a project that I was excited about. It was a really difficult decision but I just don’t have the time to make something really good and I’m so exhausted. I need a break.

We’re firmly in ‘silly-season.’ This is a period around the end of the academic year where people (including me) are short-tempered, rushed and sensitive. Students would usually start doing things like leaving laptops on busses and falling out with each other but I guess semi-online life at least avoids some of those issues. Still, a lot of the things I’ve dealt with in the last week or two have been silly-season phenomena. It’s all spurred by the approaching end of the year, the closing in deadlines and anxiety about what’s next. Like a universal cycle. Then September opens up and it all feels like spring – full of possibility and opportunity until the walls start closing in again around November and we do it all over again and again and again. Golly, you’re insightful, Revell.

Short stuff

Owen said I should watch Devs and said he will be keeping an eye on me to make sure I do. I am still rewatching Westworld and by principle will only watch one thing at once. I’ll get to it, Owen, I promise.
I twittered this paper because it’s such a succinct breakdown of the fallacies of AI research coming on the back of me finishing a lecture about hype in AI. As usual, my timing is way off. Francis did point out that although it contains references to some of the ‘greats’ in AI research it fails to mention the work of a lot of critical practitioners, challenging the story of AI.
Flight Simulator 5 continues to be interesting. As you know, it renders the world from semi-live mapping, GIS and meteorological data. So how does it treat conflict zones or areas censored in maps? Vaguely it seems.
Here I am talking about Alan’s exhibition that opens next week and I am going to see again! That’s something to be excited about.

I know these haven’t been very good recently. It is very hard to keep a balance of Excel sheets, hair-pulling and things I actually want to do and I like to think I’m pretty disciplined; that I live a life structured around targets and goals, just the incredibly tedious RPG of my life. I guess the nature of biopolitics is that you’re never good enough to please all your targets. Bring on the transhuman revolution, I say. I wanna be that goat and eat grass all day.

Love you, see you next time.

BOX037: Who am I even talking to?

28.04.2128.04.21

DS035: Get a sphere. Add some arrays in x, y and z so you have loads of spheres. Apply those arrays so they’re all objects, make sure to set the origin points to their geometry or it’ll get weird later. Make em all rigid bodies. Put a box around them and make that an animated rigid body that’s a child of your camera. Set your camera to rotate around x, y and z slowly. The box should follow it and stay locked on. Put some lights in but leave them still. Set the rigid body world speed to like 0.1 so that everything is in slow motion and doesn’t mess with your eyes. Hit bake and then animate.

The render this week is from a talk I gave for some folks last week as part of a research project called ‘Towards the Realm of Materiality.’ They’re looking at the way Philip K. Dick’s stories have been materialised so it gave me a great excuse to pick apart the Minority Report interface for an hour. It was a good event and nice to see some familiar faces as well.

I didn’t post last week because I was concentrating on writing and recording a new lecture for our Contextual and Theoretical Studies students at LCC. I’ve been struggling a bit with writing this one. The last one, which was about CGI, sort of fell together like pulled pork in reverse — the ideas, narratives and key references all appearing before my very eyes as if sent by the raster gods. This second one is about AI and imaginaries and is a bit more nebulous, it’s more like corralling balloons; I keep moving sections around and doing last minute edits.

Still from the title of the first lecture, about Computer Graphics.

If you don’t know, I’m doing three of these things, each about an hour long and have committed to making them really good. More than recordings to a slide deck I wanted to make something genuinely revisit-able and reusable. This also gives me a chance to get creative with production (which I like) and learn some more about camera work, even if it multiplies the production time about tenfold. Anyway, I’ve been crawling through the second one, reordering sections, coming to dead ends in lines of thought, reiterating bits to simplify and then re-complexify and what should have realistically taken about a week over the Easter break is still only half finished. I’ve gone a bit more ambitious this time as well, more cutaways and post-production things. I don’t know if it’s the subject or the format that’s giving me trouble really. AI and imaginaries is a bit more nebulous than CGI, it’s harder to pin down an angle for young designers to get into it from I suppose.

I also indirectly got an earful from a colleague about how shit lectures are for teaching after he saw it and some of the others being produced so that kind of put a downer on my efforts for a week or two there too. I dunno, the data for the first one is really good, most people watched the whole thing, even multiple times and I got some great direct feedback from folks. I’m not convinced that people don’t enjoy lectures – I mean I do and I can’t concentrate on anything for more than 8 minutes. Would I even call these lectures? They’re more inspired by my love of YouTube stuff really – some gags, lots of interesting examples, being energetic. I dunno. There’s a deadline of next week so I can’t put it off but I hope I end up happy with it.

Ok Revell, stop being a moany bitch, people have real problems.

We’re all doing our best

Speaking of exploring the lecture format, though I can’t put mine out because they’re for students, George Voss has put out ‘Bigger Than Before‘ recently reviewed as ‘an absolute unit of a talk’ in which she delves into scale in infrastructure and rabid blog-hounder Crystal Bennes has released the third part of ‘The Empire of Man Over Nature‘ which I’m going to watch after this.

Short stuff

I have a sort of nascent idea for another web project – when and AI is not an AI. Actually I may have written about it the last time I posted. Anyway, I dug out a thing I’ve been off-handedly referencing for ages, that a significant amount of AI startups don’t actually have any AI.
Speaking of, here’s a Devil’s Dictionary of AI terms: Big Tech’s guide to talking about AI ethics
50-ish words you can use to show that you care without incriminating yourself. Surprisingly funny for MIT Tech Review. eg. democratize (v) – To scale a technology at all costs. A justification for concentrating resources. See scale.
I know it all comes across as cynical but it is a reminder of how much guff a lot of well-meaning tech speak is and how easy it is to cut through with examples. Remember OpenAI, an organisation literally set up to ‘open’ ‘AI’ and bring public scrutiny to the work just literally sold the rights to GPT-3 the world’s biggest tech company and basically withheld it from everyone else?
The NFT bubble seems like it’s cooled down if not outright burst in some places. Everyone claims that they get it and that everyone else is missing the point which seems to be a reliable sign that there isn’t a point at all. But I liked Tokenize This from Ben Grosser which highlights one dimension of the hypocrisy.

That’s it. Once, I’ve cleared the deck on this lecture it’s back to PhD work. To be honest the lectures are all the same content as my PhD work so it’s not like it’s irrelevant, I have to read the same stuff and think about it. Shut up Revell. Ok, you know I love you, I don’t need to tell you. Be excellent; people find kindness and honesty disarming. Later.

BOX036: The Universe’s New Spin

14.04.2114.04.21

DS034: Dynamic paint but importantly looking at some generative detail; swaying boxes and texture that changes on the z-axis.

That sun is just glorious, I wish I could pull off Earth’s SS21 colour palette.

My body clock’s gone all over the place. I had the last week ‘off’ and has all sorts of plans but instead I completely crashed out and just couldn’t do anything that involved responsibility. I managed to get through about half of Disco Elysium, do a bunch of DIY and cycling and came out the other end just feeling guilty because capitalism has made me its bitch. Anyway, it’s a short one today because as usual the few hours I put out each morning to read, write and think get filled by administration as surely as nature abhors a vacuum. Most of my actual job just involves copying information from one document to another or reorganising or reorientation information to give it meaning to whoever it’s for. There’s a lot of this that goes on everyday and then I get to the end of the day and can’t think of a single thing that’s been achieved and then the guilt returns.

I suppose design is kind of information manipulation anyway. And guilt is basically a kind of high polluting fossil fuel. Anyway, I was buoyed this week by getting real deep on the results from the G-2 (that’s G minus 2, not G dash 2) experiment and the chatter around it. I may have told you this before but at school I was best at maths, physics and history and really enjoyed them but then the education system gradually beat any joy or wonder out of them by turning them into exams and all the cool, attractive people went to art school and since I didn’t have many friends except those weirdos, I followed. That’s why I bang on about history and maths so much I suppose.

Anyway, the G-2 experiment was trying to confirm a result from twenty years ago when the g-factor on muons was found to be higher than that standard model predicts. We’re talking higher by only a tiny fraction at the eight decimal place here but significant enough to be more than a statistical anomaly but not quite a discovery. G-factor is like the amount of wobble on a spinning top; as a muon spins in an electromagnetic field, the spin is titled slightly by other forces like gravity as well as the quantum foam (the particles constantly popping in and out of existence all the time). Muons are theorised to have a g-factor of 2 from gravity and then a tiny bit more to account for the quantum foam. Folks had figured that with the pretty comprehensive map of particles through the standard model that this should be cut and dry at 2.0023318362… but these folks at Fermilab have confirmed last week it’s higher at 2.0023318412… They’re going to run a bunch more tests obviously but lots of the physics people are getting bleary-eyed about ‘new physics.’ Most everything else predicted through the standard model has turned out to be bang on the money so this result indicates either a fault in the experiment (approximately 1:40,000-100,000 chance), something wrong with standard model, or extra physics that has yet to be theorised. Here’s some of my favourite YouTube channels on it: Sixty Symbols and Physics Girl.

It’s also worth noting the insane journey the G-2 went on to get to Fermilab – The Big Move.

Short Stuff

There should be a list of things that claim to be AI and aren’t AI, similar to this X Does not Exist site for GAN’s. I’ve been collecting them for my own entertainment but it feels like a lot of work to build a public repository of things that aren’t AI. Anyway, here’s another that Utah spend $20m on.
Some mea culpa on NFT’s here. I know I wrote about this the other week, but the ‘you don’t get it, you’re looking at wrong’ doesn’t change what it actually is and how it’s instantiated. Ruth Catlow at Furtherfield has also written a little push back on the anti-NFT sentiment. Furtherfield have been the leaders in critical blockchain stuff, at least in the UK for a long time and have put in years of great work so their voice is super important here and definitely holds a lot of (if not the most) weight. However, again, lambasting media for doing media and capital for doing capital doesn’t actually change what the thing is. I’m not retreading this but remember how the inventors of the Internet talked about freedom, new forms of organisation and equity of information and instead we got Google and Facebook? I think for those who are interested and engaged in the nuance of the argument (and are despairing of the humiliating gold rush) this is a useful discussion but does it actually reach the hedge fund bros and IP trolls who are making the money?
I mean this goes back to my issue with critical culture doing more to alienate the people it should be trying to affect than actually introducing new ideas. If the interest of the mainstream is ‘how can I make money as quickly as possible’ then you have to use that impetus as your lever for change rather than insisting that people are wrong about their priorities (getting rich). People don’t like being told they’re wrong.

Alright, love you very much. Oh me oh my I just want to sleep.