You know what, everyone? I…I really enjoyed having a week off. I know, I know, I’m supposed to say that I missed all the web and the links and stuff, but in actual fact it was really, really nice not to actually have to read 300 fcuking websites every day.
Yes, I know that I don’t technically *have* to do this – after all, noone is asking me (some, I am sure, would very much prefer it if I stopped) and noone is paying me and, in the main, noone cares – but the one thing that I did realise from taking a short break is that I am now in the unfortunate position where ‘writing the fcuker’ is effectively so much a part of my idea of myself at this point in time that without it I worry I might just sort of collapse into nothingness.
What I am saying, basically, is that you may want these emails to stop, *I* may want these emails to stop, but these emails may in fact never stop until I stop (in the definitive sense).
I am still Matt, this is still Web Curios, and you are probably hoping it picks up after this bit (it doesn’t, sadly).
By Shardcore
THE SECTION WHICH WITH ALL THIS SEXY POLITICAL KOMPROMAT NEWS IN THE PAPERS IS TODAY FONDLY REMEMBERING THE STORY OF FORMER LIBERAL DEMOCRAT POLITICIAN MARK OATEN, WHO WAS EXPOSED AS A COPROPHILE WHEN THE SEX WORKER HE HAD BEEN EMPLOYING TO POO ON HIM RECOGNISED HIM ON NEWSNIGHT, PT.1:
- World Sim: It always feels a bit like an admission of some sort of abject failure when I have to give the very first link of the week the whole ‘I don’t really understand what this is, but’…but, well, I don’t really understand what this is, BUT IT IS INCREDIBLE. World Sim is…it’s basically that. Imagine a text adventure in which the whole premise is ‘you’re god. On Day 0 of creation. GO!’ – well that’s what this is. Built on some LLM – I think it’s using Claude – this is one of the most amazing, dizzying, silly, brilliant applications of the tech that I’ve seen in ages – to be clear, it is utterly pointless, but it is also SO beautifully set up. Tell the programme to create humans – watch and see what sort of humans it creates. After creating the heavens and the earth, land and sea, I asked for some fauna and the system became bizarrely obsessed with developing a unicorn-based economy for my humans to exploit. Create plagues! Create existential conundra! Tell every single living creature they must engage in blood sacrifice to avoid the wrath of the capricious, all-seeing God-creature that oversees their every waking moment! I spent about an hour with this this week and I can’t stress enough how wonderful it is – I strongly advise you to just leave it in a tab and pop back in when you fancy messing with an imaginary universe at a deep, even cellular, level. Really, I think this is WONDERFUL and I hope you do too (oh, a couple of caveats – the deeper into the ‘sim’ you go, the slower it moves, so you might find that you have to reset after a while because the weight of keeping some sort of coherence makes it too slow for it to be fun – but that’s ok, because you can go back to Genesis and see what happens if you populate your sea planet with nothing but a race of hyperintelligent toasters who must fight to avoid a species-decimating rust issue. Or something).
- Vizzing The Parrot: Ok, that’s not in fact what this is called AT ALL, but I prefer my title – this is perhaps the best and clearest explainer/visualisation of How LLMs Work (Insofar As We Actually Understand That At All, Which Is Not In Fact That Far) that I have yet seen, and by far the clearest way of demonstrating to someone that, despite appearances, The Machine is literally just a probability exercise (albeit a fiendishly-complex one). Speaks the project’s creator “Using the chatgpt api, I ran the same completion prompt “Intelligence is ” hundreds of times (setting the temperature quite high, at 1.6, for more diverse responses). Given a text, a Large Language Model assigns a probability for the word (token) to come, and it just repeats this process until a completion is…well, complete.” So what this does is show you ALL the branching options that the model has considered to complete the sentence – and a picture of the way in which it conceives of the latent space within which the concepts exist, and how the relational structure between said concepts works (to an extent). This is obviously limited – I am pretty sure that one of the curiosities of latent space is that it’s not really possible to viz it in a way that accurately captures all the relational data, or at least not in a way that we could reasonably be expected to make sense of – but it’s SO interesting, and there’s something dizzying about the way the visualisation shows you the sheer breadth of potential options and directions in which the ‘thinking’ The Machine is doing can go. Honestly, if you’ve struggled to get your head round ‘what is actually happening when an LLM is doing that thing it does’ then this really might help.
- Terra: Oh, I really like this and I sort-of want one. Terra is an open-source project which anyone with a 3d printer and a reasonable amount of technical nous (so, in all likelihood, approximately three of you) can participate in, which basically lets you create an AI-powered GPS pebble that will take you on MYSTERY WALKS. The way it works is that there’s a GPS tracker in the pebble, which gets programmed by an AI based on the prompt you give it for the sort of walk you want to take – it knows where you are when you make the request, and based on the prompt you give it (eg “a two-hour stroll around the most picturesque areas of Catford”) (for non-Londoners: this is a gentle gag) it will create a route for you that will take you, eventually, right back to where you started. It’s then your job to use the pebble to guide you around the selected route – all it shows you is a compass direction to walk in, with no screen or additional info, with the idea being that you just follow the broad indications of where to head and enjoy your stroll without distraction, safe in the knowledge that you’ll eventually end up right where you started. I LOVE THIS – as someone whose idea of ‘a good day out’ is often ‘leaving my house at 10am and basically just walking until I feel like I might die’ it’s pretty much perfect for me, although I accept that there is a probably-not-entirely-safe-at-this-exact-point-in-time degree of trust being placed in the black box of AI here and I wouldn’t necessarily trust this to send me on walks anywhere I wasn’t already a BIT familiar with (I wouldn’t, for the sake of argument, suggest using this as a way of ‘taking my first wander around Cartagena at night’), but it’s SO LOVELY. I want one. Can, er, one of the three of I alluded to earlier sort this out for me? Come on, I never ask ANYTHING of you (other than, er, the time it takes to read 10,000 words a week of this sh1t. On which note, ffs Matt! Come on! Links!).
- Cursor Watching: I am of the personal belief that ‘multiplayer websites’ are going to become a bit of a thing in the not-too-distant future (file this under “stuff you can remind Matt of in the future to remind him he’s fcuking terrible at making predictions and has repeatedly promised that he will stop making them”), and this website collects a nice collection of examples of the genre which illustrate how FUN they can be. Click the link and you’ll be taken to a page on which are embedded windows onto a bunch of other sites, each of which is a space which plays with the idea of multiple users, all strangers to each other, experiencing it at the same time. So you have sites where you can all see each other’s cursors, letting you interact through gestures with strangers, or ones in which you can make collaborative music…some of these have been featured in Curios over the years, but there’s a lovely selection here which will be useful for you all when you FINALLY do what I have been asking you to do for years and ensure that every single web project you’re involved in includes a fun element like these because WHY NOT?
- SearchMySite: A project seeking to create a small, curated search engine which focuses on sites from ‘the indieweb / the small web / digital gardens’ – this has been running for a few years now, apparently, but I only came across it this week, and from what I can tell there are a reasonable number of sites that it crawls when pulling results. You can submit a site for inclusion in the corpus (although I couldn’t personally make the form work when I tried again just now), and, even if you don’t want to use it for search, I can highly recommend spamming the ‘random’ button and seeing where you end up – I obviously can’t vouch for everything on here, but all the sites I’ve found through it so far have been…nice, and, crucially, avowedly non-commercial.
- Melon King: I am quite angry that I don’t seem to have stumbled across this before now, because, let me be clear, IT IS EVERYTHING I HAVE EVER WANTED FROM THE WEB (oh, hang on – I think I have featured this page before, during COVID). From the ‘about’ section: “This site has been under construction since 2016; it is not a nostalgia site, and its not entirely a personal site – Melonking.Net contains a tangle of biographical material, fiction, idealism and memories. It was created to be the website I imagined creating when I was 6 years old; a recollection of the web through the infinity mirror of time. This site is about the web of media and life; it beckons you to explore, there is no algorithmic guard rail; sometimes its serious and often its silly, but its always genuine – as you explore I hope you’ll let go of the world you are in, and inhabit this one for just a while :^]” This is…this is dizzying and mad and a sort of odd, half-remembered fever-dream of the internet in 2000-ish, with mad graphics and an insane hodgepodge of interfaces and 3d animations and glittery stickers and GeoCities vibes and some actually pretty impressive webwork going on under the hood, and SO MUCH STUFF – so many rabbitholes and easter eggs, all of which is also, at heart, an intensely personal account of someone’s interests and life, on and offline, and I can’t stress enough how utterly, magically wonderful this is. I think I have said on here before (SO MANY TIMES YOU ARE SO FCUKING BORING MATT) that the best websites are the ones that basically feel like you’re stepping into an ornate and beautifully-rendered extension of someone else’s brain, and this is EXACTLY how this feels. I love it and in all honesty I would happily sack this off for the morning and spend the rest of the day with the Melon King (but you’re not getting off that lightly).
- Favoree: This self-describes as ‘IMDB, but for YouTubers’ – not an influencer discovery platform, though you could probably use it as one if you really tried hard, but instead a way of collating YouTubers around particular topics of interest, offering information about the topics they cover and the style they do it in, and a means of finding new streamers you might be interested in, away from the algorithmic pressure to watch some shouting lunatic whose got an inexplicably-huge subscriber base. I can’t speak to how useful this is being a non-YouTube person, in the main, but here it is just in case.
- The Pleasure Gallery: ACTUAL BONGO ALERT! Textual bongo, to be clear, so as long as noone’s actually reading what’s on your screen you can click with abandon here. The premise here is simple – artist Angela Vang made this site to share a random selection of sexts sent or received by her and her friends (or anyone on the web who submits one via the site), which are presented as a scrolling, seemingly-infinite conversation between two nameless, genderless and EXTREMELY HORNY parties. The decontextualisation of the messages and the lack of continuity as to the ‘who’ on each side makes this pleasingly abstract whilst at the same time being VERY FILTHY INDEED. No judgement whatsoever, but there are a LOT of people who like rimming featured in these messages. I love this, and it feels like a ‘proper’ gallery piece.
- Mosaic: This feels like a potentially-interesting idea, but I am somewhat-sketchy on how it ACTUALLY works – the premise behind Mosaic is that there should be some sort of database of digital creatives’ work, a record of what has been made by whom that can be referenced and referred back to as ‘proof of creative ownership’, and that’s the function the site’s hoping to fulfil. The idea is that ‘creators’ make an account, and for each project they complete they create a ‘block’ on Mosaic which acts as a record of their creative ownership – per their description, it’s “an innovative online platform designed to help you showcase your work accomplishments in a more detailed and authenticated way. Think of it as an interactive, project-based resume that goes beyond just listing your roles and responsibilities. On Mosaic, your work is represented through ‘Blocks’, which are individual projects you’ve worked on. Each Block contains specific information about the project, such as what the project was, who it was for, when it was completed, what role you played, and who you collaborated with. But here’s the exciting part: these Blocks aren’t just statements, they’re authenticated. That means when you create a Block, you send it for approval to someone who can verify your work, adding credibility to your professional achievements.” Now, I don’t know about you but I hear the words ‘block’ and ‘authenticated’ and I immediately start to get the web3cryptoick…but it’s totally unclear whether this is an unfortunate coincidence, or whether this is blockchain-linked but just trying to hide it and look normal. Anyway, this feels like a potentially good idea which is probably never going to achieve meaningful scale (SORRY GUYS I HOPE I AM WRONG), but, equally, it probably can’t hurt to check it out if you Make Stuff On The Web.
- Food Mood: New fun Google AI experiment thingy! Food mood is a silly – but weirdly-compelling, to me at least – little toy which lets you pick two cuisines, a type of meal (starter, soup, main, dessert – it’s hardly Larousse, but wevs) and a number of diners and LIKE MAGIC it will imagine a new, frankenfood hybrid of the two cultures complete with a recipe, ingredient quantities and a photo of the resulting dish. To give you an idea of the sorts of outputs, I asked it for an Anglo-Italian hybrid starter and it offered me a ‘recipe’ for a ricotta, smoked salmon, olive and caper bruschetta, a combination so violently-unpalatable to me that I had to go and mentally wash my mouth out with Cynar, but your mileage may vary – if nothing else, if you and your partner are from different cultures you could reasonably let The Machine plan out the most romantic, culturally-confused and inedible date night supper you’ve ever had.
- Musical Canvas: Another fun Google AI experiment thingy! Musical canvas is a nice little bit of multimodal experimentation – doodle whatever you like on the canvas (the prompt says ‘draw’, but has anyone ever been able to actually ‘draw’ anything using their mouse? I posit that they have in fact not) and The Machine will attempt to interpret your daubings and render them as an appropriate musical output. The resulting songs are trash, to be clear, but it’s fun to draw a succession of increasingly-obscene cartoon phalli and watch Google desperately attempt to reinterpret them as ‘a poorly-drawn ten-gallon hat’ or something and spin up a jaunty banjo-led country instrumental to accompany it.
- Instrument Playground:ANOTHER fun Google AI experiment thing! This is surprisingly deep – at first glance it’s a reasonably-simple music generation toy – select your instrument, give it some vague directional prompting to fix a style and away you go – but it turns out you can actually noodle with the resulting sonic outputs in a range of ways, manipulating individual bits of the 20s track which gives an interesting top-level overview of the way in which generative AI can transform the process of editing digital media. I am basically tone-deaf and so get limited joy from this, but the more musical among you might find it a fun thing to play with. All these Google bits came via Lynn Cherny’s excellent Things I Think Are Interesting newsletter, by the way, which is a gem if you’re interested in new AI developments.
- April Cools Club: A simple premise, started a few years ago: “The idea is pretty simple: on April Fools’ Day (also known as “April 1st”), a participant produces genuine content that’s very different from their normal produced content. It could be a different format, a different topic, a different style, anything. The constraints are: It is something they normally wouldn’t do. It is totally genuine: no irony to it.It is up to their usual standards of quality.For example, some might normally post complex software engineering content to their blog. But this April Fools’ Day, they are publishing an essay on microscopy, how they got into it, and what it means to them, complete with a gallery of their favorite microscopy photos.” This is a collection of the various projects various people have undertaken in Aprils past, and a place to see what people are doing this year, and while there’s nothing particularly remarkable happening here I personally find something genuinely pleasing about stuff like this and like to celebrate it where I can (aren’t I nice? Jesus, get OVER yourself you twat).
- Piet Dewijngaert: This is the personal website of one Piet Dewijngaert, a(n, I think) Dutch developer who’s built this to showcase his digital design and build skills and it is VERY SHINY – in particular the ‘captcha’ elements are really beautifully done, lovely interaction design here. Very nice work indeed.
- Global Outdoor Days: Well this is some sobering data. Global Outdoor Days lets you pick any one of a number of countries, set a minimum and maximum temperature range during which you consider it personally comfortable to be outside, and then click to see what sort of projected difference climate change is likely to make over the coming decades – I tried it with Rome just now, suggesting that I would prefer not to spend too much time outside when the mercury tips 33 degrees, and learned there are likely to be 54 fewer days when I will want to go outside by 2050 (I will be dead by then, no question, but the point holds). The UK? 64, apparently. GOOD LUCK EVERYONE.
- Chatbot Arena: This is an interesting ongoing experiment – Chatbot Arena is working to compare the outputs of different LLM models based on people’s qualitative feedback – feed it a prompt and it will spit out responses from two different models without telling you which ones they are. The users then selects which they think is the ‘best’ answer, all of which data contributes to a live league table of ‘which model is best right now’ – interestingly (I mean, not THAT interesting, fine) Claude’s Opus model is currently narrowly edging out GPT4, with everything else a significant way behind.
- All The Floyd Animations: This is quite amazing. I have never really been a fan of 60s/70s music and as such the extreme degree of reverence granted to bands like Pink Floyd has always felt slightly baffling to me – stuff like this, though, hammers home just how INSANELY popular they still are. The band recently ran a contest asking people to create animations to accompany the songs on the band’s Dark Side of the Moon album, as part of an anniversary rerelease – MY GOD did the fans oblige. This YouTube channel collects ALL the videos submitted as part of the contest, and fcuking hell is there some range here, from things that could best be described as…er…’charming examples of outsider art by some very committed amateurs’ to genuinely-Oscars-level claymation work. Seriously, just have a bit of a dig through these, it’s a quite astonishing kaleidoscope of styles and skills (but you will get REALLY fcuking sick of Pink Floyd after about three of them, if you’re me at least).
- One Minute Park: I LOVE THIS PROJECT. One Minute Park is a very simple website/project – click the link and you get presented with a full-screen video, in landscape, which lasts for exactly 60 seconds and which presents a delightfully, perfectly mundane scene filmed in a local park somewhere in the world. That’s it – after 60 seconds, the scene shifts to another park somewhere else on the planet. No more, no less, just small, evanescent windows into slices of quiet humanity. Anyone can submit a video, and I think this is a lovely way to spend a minute this weekend should it ever stop fcuking raining.
- After The Beep: I am always a sucker for ‘here’s an answerphone, leave a message to the entire world’ websites, and this is exactly that – the number’s in North America and I don’t think, based on listening to a few of these, that there’s any moderation going on, and it’s weird and profane and silly and lonely and sad and feels oddly like the past rather than the future, and I could listen to these all day. Occasionally VERY NSFW, just so you’re aware.
OUR NEXT MIX IS ANOTHER BUNCH OF BEATS AND BLEEPS MIXED BY FORMER EDITOR PAUL!
THE SECTION WHICH WITH ALL THIS SEXY POLITICAL KOMPROMAT NEWS IN THE PAPERS IS TODAY FONDLY REMEMBERING THE STORY OF FORMER LIBERAL DEMOCRAT POLITICIAN MARK OATEN, WHO WAS EXPOSED AS A COPROPHILE WHEN THE SEX WORKER HE HAD BEEN EMPLOYING TO POO ON HIM RECOGNISED HIM ON NEWSNIGHT, PT.2:
- Geddit: I genuinely had no idea that Reddit was used by some people as a sort of online shopping destination, sort of like Facebook Marketplace for people who unironically use the word ‘normies’ in actual, out-loud conversation, but apparently it is and Geddit is a nice little window into the slightly-odd world that is peer to peer online selling. You can filter by location or specific subReddit, but there doesn’t seem much going on in the UK – instead, it’s fun just to scroll through and treat it as a sort of temperature reading of Middle American wants, by which token WOW are people worryingly into flickknives in the States right now!
- Cellular Automata: I genuinely don’t understand what this is or what it’s for, but it makes oddly-pleasing low-res black and white abstract pixel renders which for some reason I found oddly soothing and for some reason you might too.
- The Curricula: This doesn’t feel like a good idea, I must say. “What if Wikipedia, but spun up entirely by an LLM?” is the question which as far as I can tell the Curricula appears to be trying to answer – type in…well, anything related to an area of study or field of knowledge and the site will spit out a bunch of ‘information’ (or at least, information-shaped words) and a bunch of related concepts you can ‘dive into’…except it’s all so depressingly *thin* and *stupid*, like the very worst of tissue-thin AI-generated prose, and it feels like gruel by comparison to the FILLING, HUMAN-COOKED MEAL that is Wikipedia (this is both a terrible metaphor and a weirdly-funny one – imagine going back 20 years and suggesting to anyone that Wikipedia would one day be looked upon as a heartwarming example of collective human effort and endeavour!). If you want a vision of the level of ‘content’ that we’re going to get as a result of the ubiquity of LLMs, this is very much it and it is empty and stupid and hollow.
- Am I Flying On A Boeing?: Plug in your flight number and let this site tell you whether you need to supplement your aviophobia with, er, Boeingphobia! Feels like a rival plane manufacturer could reasonably chuck whoever’s behind this a few quid for a badging exercise here.
- Palmsy: Have you been enjoying the latest smartphone-related moral panic? No, me neither tbh. Still, you probably have OPINIONS about the extent to which ether smartphones have ruined lives OR the problem is and always has been PEOPLE and SOCIETY and PROBABLY CAPITALISM (delete per your personal preference) – Palmsy is either a darkly-funny bit of satire or a slightly-troubling reflection of how broken we all are by Posters’ Disease (probably a bit of both tbh), which presents itself as a new social media app with all the functionality you’d expect, posting and liking and sharing, except…there’s no social network there, it’s all a ‘game’, and all your posts exist only within the app; the ‘reactions’ and ‘responses’ you get are all from the app itself, with commenter names ascribed to real people from your contacts so you get the illusion of a real network…IS IT ART? IS IT DIGITAL METHADONE? I genuinely have no idea, but I found some of the comments on the App Store listing a bit…weird, tbh. “Palmsy is just fun. The people “liking” your posts didn’t actually see your post. Palmsy just pretends they did by pulling random folks from your contacts. It’s funny to see who is still in my contacts. Then, I weirdly get the dopamine hit of social media likes without the indignity of posting on social.” That’s the interesting bit, I think – the fact that the dopamine hit exists with nothing behind it, which admittedly sounds like there might be something psychologically…a bit *off* with our relationship with this stuff (STOP THE FCUKING PRESSES, I KNOW!).
- NewspaperMap: ALL OF THE WORLDS NEWSPAPERS, ON A MAP! Ok, fine, I have no way of knowing whether it is in fact ALL of the world’s newspapers, and I suspect it probably isn’t, but there are definitely a LOT – zoom around and see where their offices are and get links to their websites. Surprisingly useful if you want to get an overview of local media in a specific geography, and, presumably, should you want to finally make a pilgrimage to the Socialist Worker HQ (it’s in Deptford. Of course it’s in Deptford).
- A Call For Crab Jokes: I had no idea that there was a Crab Museum in Margate (I suppose there needs to be more than just the art gallery and half-a-dozen terrible watercolourists to sustain all those chronically-overpriced small plates restaurants), but there is, and they need YOUR help. 26th of April is INTERNATIONAL CRAB DAY (I am telling you in plenty of time so that you can start planning the street party), and the museum is celebrating by (and this is slightly baffling to me, but wevs) asking YOU, the great unwashed, to submit their entries for THE BEST JOKE ABOUT CRABS OF ALL TIME, which will be judged by a genuinely-impressive panel of comedians to find the winner, which will be honoured at a special ceremony. I’ve spent the two minutes it’s taken me to type this bit trying to think of a crab-related joke and I am fcuking BAFFLED, but then again I am approximately as funny as AIDS and I expect you all to do significantly better. Should one of you win, I ask only that you dedicate the victory to ME.
- Get A New Emoji Approved: It’s that time again! The Unicode Consortium, the oddly-shadowy body that oversees the emoji universe (they’re not shadowy, to be clear, just that the name very much *sounds* like they should be – protip, if you want to sound benign and not in fact massively sinister, maybe avoid the word ‘consortium’), is once again accepting suggestions for new emoji that they should add to the canon (over time – this is a slooooooow process). I have featured this before in Curios, because I think it’s a genuinely interesting and perfectly-democratic opportunity to actually have a lasting, if small, impact on human culture – I fcuking HATED emoji when they first became part of our common lingua franca (look, I like letters and words, I am a traditinionalist) and I still largely disdain them for personal use, but I grudgingly accept that mine is a minority viewpoint and that they are hugely useful for billions, and they are culturally significant in a way that my having spaffed out over a million words of this sh1t over the past 13 years will never be (not bitter), and the idea that YOU, gentle reader, could be responsible for the creation of an entirely new communicatory symbol that will be used by people who haven’t even been born yet is quite magical, even for dead inside me. You can read a decent guide to ‘what they are looking for in a new emoji’ here – GO, WEBMONGS, GO!
- British Place Name Finder: Would you like a website which will help you find every single village in the UK with the word ‘bum’ in its name? Yes, yes you would, don’t pretend that you’re ‘better’ than that. This is potentially really useful for anyone searching for places with specific name-types for PR stunt purposes, but also for really, really puerile reasons. Why not use it to plan a ‘childishly-named places’ holiday tour of the UK?
- Fightback NYC: This is interesting – not sure what I make of it, but it’s certainly a curious idea. The blurb is as follows: “Fight Back is a theatrical experiment. It’s March 13, 1989, and you are attending a meeting of ACT UP New York, the passionate group taking direct confrontational action to fight the AIDS crisis. But here’s the thing. There are no actors. It’s not a performance that you sit and watch. You and everyone there are active participants in the meeting.” It’s basically ‘Historic Moments In Social Justice: The LARP’, which, I’ll be honest, makes me do some pretty intense internal self-pretzeling at the imaginary awkwardness of it all, but this is taking place in NYC and hence I imagine all the people participating will have that peculiarly North American level of self-possession that means that they can do sh1t like this and not feel like a terrible fcuking pseud. Anyway, should any of you reading in New York happen to check this out I would love to know what it’s like to experience.
- OffShore Wind: A global map of offshore windfarms, presented more out of curiosity than because I expect any of you to do anything with it. It did strike me, though, that there really should be more of these and I was slightly surprised by how sparse some regions are in terms of coverage (although, equally, I am utterly ignorant of the meteorological conditions needed for a wind farm to be viable, other than the very basic ‘wind’, and as such should probably shut up).
- Not To Do: More app-based artfun from Friend of Curios Damjanski – Not To Do is a very simple app with a very simple premise, which is that each day it will present you with a list of three tasks that you definitely, 100% SHOULD NOT DO that day. It’s $3, because artists have to eat, but as far as pointless, silly, funny apps go I rather like it.
- Ahead: An old art project that I have only just come across: “In English, AHEAD means to move forward, to lead or progress. The title is also a pun: A HEAD (one head). Kruithof began this project by thinking about how to create an anonymous portrait, where the subject’s identity remains private. By capturing the back of the head, one cannot recognize gender, nationality, age, and emotions of the sitter. Removing these features, which always fall within the tradition of portrait photography, unifies all resulting pictures. In addition, facial recognition algorithms are unable to identify or verify a person’s identity from a picture like this: AHEAD shows a flaw in human encyclopedic tendency by means of anti-labeling and anti-classification. For AHEAD, Kruithof took 1,080 pictures in Berlin, New York and Mexico City using an iPhone. She left each subject to choose her or his background color, as if they were taking a selfie, but facing the background instead of posing in front of it. For the final installation, the pictures have been arranged in a grid, the same way digital photographs are usually organized online and visualized on mobile devices. From the distance, each portrait in this series looks like a single dot, resembling a pixel forming, in turn, a bigger picture.” There’s something very pleasing indeed about the accumulated backs-of-heads here, though I couldn’t quite explain why.
- Latent Places: Via Jamie Stantonian comes this excellent TikTok account which shares AI-generated horrors which have been treated with some pleasing filters to give them the grainy air of handheld camera footage, along with 50s-doc-style AI voices, making the weird all the weirder. Really nicely done with a very strong unifying aesthetic (/pseud).
- Ways of Seeing: This is SUCH AN INCREDIBLE INTRODUCTION TO ART! Classical art, to be clear, but for anyone interested in art history who wants a ‘from the top’ grounding, this is superb and so, so interesting. “Based on the 1972 BBC series and comprised of 7 essays, 3 of which are entirely pictoral, Ways of Seeing is a seminal work which examines how we view art” – honestly, this is really wonderful, a proper grounding in the social and economic and cultural backdrop to works across styles and eras. In 2024 this would be a 9h YouTube video – which is fine, but sometimes words are good too.
- A Green Fuzzy Ball: In your browser. Pick it up. Move it around. Marvel at the physics of the little green hairs. That’s it, we can move on now, but wasn’t that nice?
- Concrete: The world of specific and niche professional organisations always intrigues me (I think it stems from discovering that there is, amongst the Guilds of the City of London, a Worshipful Guild of Basketmakers, who I can’t imagine are THAT busy at this particular juncture and who must feel a bit…inferior when hanging out with, say, the goldsmiths or the bankers), and so I was thrilled to discover that not only is there a Global Cement and Concrete Association (there is!) but also that they run an annual photo contest (which must at this point be the only fcuking annual photo contest I have yet to feature in this sodding newsletter) celebrating the very best photos of, er, stuff made out of cement and concrete. Ignore the environmental impact of the building materials in question (no, really, you have to because otherwise you can’t really enjoy them) and instead enjoy the ENORMOUS GREY EXPANSES on view here – not all of the images are stellar, to my mind, but there is some quite remarkable architecture on display in some of these, and the framing in a few shots is exquisite. One small note – the Global Cement and Concrete Association doesn’t appear to have a mascot, which feels like a MASSIVE OVERSIGHT and an excellent opportunity if you ask me (“Conky The Mixer”, anyone?).
- The World Nature Photography Awards 2024: From brutalism to beautiful nature, now (SEAMLESS, Matt, seamless!), with this year’s World Nature Photography Awards – I say this every year, I know, but personally I find all these images so post-produced and HDR’d that I don’t really see them anymore, but you may have a higher tolerance for this particular style than I do. You can buy all of the featured images as prints, should you so desire – which does rather beg the question ‘who the fcuk is buying a wall print of that photo of all the severed monkey heads in a basket, because fcuk me that is some bleak domestic art?’.
- Flipbook: The Dataviz wizards at The Pudding are running a little experiment in collaborative webwork – they’re asking people to submit a single hand-drawn frame to contribute to the longest collaboratively-created flipbook animation ever. Click the link, draw a frame, sign up to get alerted when (if) it ever gets completed. This is, per their explanation, a question about data corruption: “This is essentially an experiment in generation loss; how will the original drawing mutate as it is traced by more and more people?” Genuinely curious to see the end result of this.
- Insane Interior DesignTok: You know people say that we haven’t quite worked out how odd ‘content’ is going to get in the very imminent future? I felt that when watching these. This TikTok account shares CG videos of VERY VERY ODD interior design stuff, with an equally odd, stilted, AI-generated voice-over…I don’t quite understand the workflow here or how these are being made (the CG feels ‘proper’, though), but there’s something undeniably, weirdly compelling about them…I think the weird thing is that I DON’T UNDERSTAND WHY. It’s scratching something somewhere in my brain that I recognise but can’t quite define, and I wonder whether this is something we’re going to get more of – stuff that The Machine works out ‘works’ for people, and which appeals to us but for reasons that we simply can’t define because it’s exploiting patterns or quirks that we ourselves haven’t noticed yet…Anyway, this is fcuking weird and I like it a lot – no idea how much if anything is AI’d here, but it FEELS of that ilk of thing.
- Guess My RGB: Move the sliders until you’ve matched the RGB value to the background colour on-screen. Which I appreciate sounds about as much fun as self-administered urethral surgery, but which I promise is significantly better than that!
- The New York Times Simulator: A wonderful little game by the perennially-talented Molleindustria, who’s created this game in which you have to ‘run’ the New York Times, selecting headlines for the front page which will appease the various audiences you’re beholden to. This is SATIRE, and so the stakeholders in question are the Police, the Rich, and Israel (on which point, regardless of your position on the conflict in Gaza right now I don’t think it’s possible to look at the way it’s been reported in the NYT and not think ‘yes, ok, there is possibly some pandering here’), and there are some nice touches such as the ability to rewrite headlines into the passive voice to curry favour.
- Knot: Our final miscellaneous link this week is this very simple game which sees you attempt to unscramble the links by moving tiles around – this is SUPER GENTLE, and scratches a very particular part of my brain that doesn’t get scratched as often as I feel it might want to be.
OUR LAST MIX THIS WEEK IS THIS SUPER 80s-ISH LASERDISCO-Y SELECTION BY PRIZ!
THE CIRCUS OF TUMBLRS!
- Falling Down The Internet Hole: SO MANY EXCELLENT LINKS! “Falling Down the Internet Hole is a side project of lagazettedumauvaisgout.com. Falling Down the Internet Hole is a website database selected on the below criteria: old-school design, bad taste, kitschy and/or weird.” All of this – honestly, so many of these were new to me and it was honestly tempting just to keep this to myself and use it to source future Curios – SEE HOW UNCOMMONLY UNSELFISH I AM?
THE TROUGH OF (INSTA) FEEDS!
- Patchwork Letters: Sent to me by Kris, this is a lovely Insta maintained by one Alicia, in which they share pottery and paintings they are making – there’s an occasional series of little watercolour squares of that day’s sky which are genuinely beautiful.
- Parisian Chairs: Literally just a collection of photos of the patterns on chairs in Paris – you know, the ones whose seat and back are made from woven plastic strands in different colours? Yeah, those ones. Like all these sorts of collections, this is unexpectedly and slightly-weirdly compelling.
LONG THINGS WHICH ARE LONG!
- Dopamine, Desire and the Internet: We kick off the longreads with a chewy one, a piece which feels very Of The Now what with all the DISCOURSE about smartphones and social and What They Are Doing To Us All, and the wider conversations about media and consumption, and the chatter that’s still chuntering on, motivated by Ted Gioia’s ‘dopamine culture’ piece that did the rounds a month or so back. This is a really smart and well-thought piece which exists slightly in opposition to Gioia’s take – LM Sacasas argues persuasively that it’s not dopamine rewards which are driving this fragmented, fast-moving and seemingly-superficial cultural moment, and while you really do need to read the whole piece (it is worth it, I promise) you can get a sense of his arguments from this para: “The organizing principle of this essay has been this: the “dopamine culture” frame is too simplistic and tacitly encourages an impoverished view of human personhood. To reduce a discussion of this significance to the operations of dopamine already sets us off on the wrong path. We need a fuller account of our relationship with digital media as well as a richer story of human desire in order to see our way through the challenges we face. Interestingly, the dopamine framing is also an artifact of the condition it tries to explain: it is a powerful and catchy meme, although one that is offered in the best spirit. For these reasons, I fear that it may trap us in the very patterns that it seeks to overcome.”
- 14 Years Of Tory Rule: This is a couple of weeks old now, but if you haven’t read it yet then you really, really should (sorry – this applies to people living in the UK; the rest of you, feel free to ignore our pathetic parochial issues and focus on your own, of which I am sure there are myriad) – in the New Yorker, Sam Knight gives what I think is the best precis of what the past decade and a half of Conservative administration have wrought in the UK, and if you can read this without becoming infuriated by both the actions taken and the STAGGERINGLY unrepentant quotes from the principle actors involved then, well, you’re a stronger person than I am. This is, above all, a story about stupidity, arrogance and venality – qualities which any good Brit will know have long been hallmarks of the core Conservative voter! I have a good friend who will occasionally get into quite impassioned rants about how the two worst human beings since Hitler have been Donald Rumsfeld and George Osborne, and on the latter I have to say this piece has convinced me.
- Why Capitalism Won’t Save The Planet: I know – shocking, right? This is actually a review of a book of the same name in the LRB, written by William Davis, which neatly and cogently sets out all the reasons why it’s increasingly silly to assume that we can capitalism our way out of this environmental pickle we seem to have somehow landed ourselves in. This clear, sobering and not a little depressing, but the main point to take away from this – quite bored of this one, really, but it seems to bear repeating – is that FINANCE IS THE PROBLEM HERE, and that when you have a system which is effectively underpinned by speculation you are inevitably not going to get ‘the market’ providing the incentives that you need. “Oh no, it turns out the hedge fund and the pension fund were more concerned with immediate predictable profits than whether or not the pursuit of said profits might make the planet unlivable within a century!” – WELL IMAGINE THE SURPRISE. This is not particularly cheering, but might radicalise a bit further which, honestly, wouldn’t be a bad thing.
- The Deaths of Effective Altruism: This overview of the movement in WIRED is a really good one – it does an excellent job, better than I’ve seen in most of these sorts of pieces, of drawing an intellectual line between the more traditional utilitarian questions raised in the mid-20th-C and the rather more…out there stuff espoused by the EA/Accelerationist movements in the past couple of years, and made concrete something I’ve struggled to articulate previously when thinking about this stuff – to whit, that they problem with EA stuff is that it places so much weight on data and modeling that data and modeling can’t really support, and that it falls down where this stuff ALWAYS falls down, since JSM, when it comes to attempting to ‘quantify goods’ in any meaningful, sensible way that doesn’t lead you into some genuinely appalling cul-de-sacs. Anyway, this is a good, approachable and interesting exploration of one of the rare instance of philosophy going almost-mainstream – worth a read, even if you wouldn’t ordinarily have any truck with the ‘p’ word.
- Some Thoughts On Online Culture: Katherine Dee writes a newsletter called Default Friend which I started subscribing to recently – this is a post in which she tries to articulate what the newsletter is about, which doubles as a series of really interesting observations about How Culture Works Online and How It Intersects With Society (no, come back, wait!). It opens like this, which should give you a decent idea of whether you want to read the rest or not – but if you work in branding, strategy or ‘culture’ or planning then I humbly submit it might be a useful series of principles to consider, eg: “Everything is downstream of fandom. Both in the sense that many of the quirks of modern culture are a direct result of the influence of media fandoms and that, overwhelmingly, the way we organize ourselves is fandom. So not only are most political movements essentially fandoms, but pre-existing media fandoms, like the ones we saw on Tumblr in the 2010s, impacted how we talk and think about ourselves. This started in the 1970s and became visible to everyone in 2015.”
- Models All The Way: A SUPERB explainer all about how AI models are trained – talking you through the datasets used and what the machine does with them as it ‘learns’. So so so so interesting, pleasingly visual in execution, and a nice companion piece to the link I posted all those hours ago (3h20m of typing, to get from there to here, in case anyone’s wondering – no, I know you’re not, because who wants to know how the Curios sausage is made? NO FCUKER, etc) about ‘how LLMs ‘think’’, and why they are biased, and where those biases come from. Sort-of an essential, basic level of understanding of ‘how this stuff works’ is, I think, really important, and stuff like this is vital in developing that.
- Political Bias In AI: A really interesting piece in the New York Times, looking at the political orientations of different LLMs based on answers each gives to the Political Compass quiz – this shows a notable ‘liberal’ bias in most models, although I question how useful this sort of split is to anyone outside of North America (and even there, to be honest), but it’s more interesting when it gets into how easy it is to shift these with training, and when you start to think about the whole ‘magical friend in your phone who tells you what to think and do’ element of the coming future it’s this bit that feels properly interesting. Ultimately this is, as the piece sort-of acknowledges halfway through, less about LLMs and more about the way in which the Overton window, particularly in the US, has gone slightly mad in the past decade.
- Anthropomorphising The Machine: Curios favourite Ethan Mollick has a book about AI out this week, and to accompany it he wrote this post about why you get better results from LLMs when you talk to them like a person – as ever it’s a useful read, and it reminded me of a slightly unpleasant meeting I had recently in which I blithely dismissed using GPT4 for mundane work tasks as being like ‘dealing with all the stupid people in a PR company, but significantly less irritating’ and WOW did that get frosty quickly.
- Microservitude: Jared Shurin’s newsletter last week looked at the legacy of Microserfs by Douglas Coupland, and the extent to which the personas described in the book – intelligent, introverted, arrogant – have in many ways been those that have shaped significant swathes of digital, and therefore actual, culture in the decades since the book’s publication. As Shurin writes, “In both Microserfs and reality, tech’s corporate culture has come to encourage that division. Why foster a moderate belief that, actually, there are all sorts of lives and normalities that can mix in broader society when instead you can recruit ‘unappreciated geniuses’ by playing to their insecurities? Come to an exclusive place that understands how very special you are. We’ll even do your laundry. Microserfs, with its cast of coders obsessed with flat foods and blind worship of ‘the Bill’, paints a faintly tragic picture. But thirty years on, it now feels a lot less quaint, and a lot more frightening.” I enjoyed this a lot, in part because of my age – Microserfs was a canonical text – but also because I don’t think Coupland’s work gets referenced enough in current times (yes, I know he named a generation, but I think the relevance of a lot of the cultural commentary in his books has been lost in the post-digital era).
- Fake Email Jobs: Do you want to know why I think we’re heading for a jobpocalypse? Stuff like this. Do you know how many people do jobs which are literally just ‘taking information, changing its shape and putting it somewhere else’? Do you know what LLMs are particularly good at? Well, yes, exactly.
- So You Think You’ve Been Gaslit: I’ll be honest, I am including this mainly because ‘gaslit’ and ‘gaslighting’ rival ‘woke’ for the title of ‘word with most wildly fluctuating set of meanings currently in use today’ (and are also increasingly edging into the category, also alongside ‘woke’, of ‘words which if you use them will make me think marginally less of you than I did previously’) – this New Yorker piece looks at the term’s phenomenal rise to prominence, and the different ways in which it’s being used, and the extent to which ‘they were gaslighting me’ has come to mean, variously, ‘doing stuff I don’t like’ or ‘lying to me’ or, occasionally, ‘doing some incredibly dark and psychologically appalling sh1t’.
- Secrets of Japanese Urbanism: I’m not a particular Japanophile, but I’ve been interested enough in the country and its culture to have had a vague idea of the particular idiosyncrasies of Japanese cities and their layout and architecture, in part through reading loads of novels set there and in part through the writing of Craig Mod, whose walks through Tokyo and elsewhere I’ve featured in Curios before and who always waxes rhapsodic, with good reason, about the unique character and human-ness (for want of a better term) of urban spaces in the country. This is a really interesting article which explains the political and economic reasons behind why the country’s cities have evolved as they do – I don’t mean to turn this edition of Curios into some sort of CAPITALISM=BAD! screed, honest, but it does rather feel that part of this is down to ‘stuff being given the opportunity to just exist as a small concern without being forced to become massive just to survive’.
- Argentina and Lithium: Oh, fcukit, no, it IS another ‘capitalism=bad!’ screed! Ok, this is a slightly different angle – this piece is about Argentina’s access to lithium deposits, and the decision taken by recently-elected pseudo-libertarian madman Javier Milais to offer access to said deposits up to the highest bidder rather than using the country’s reserves for the benefit of the national populace through a previously-planned scheme which would have augmented the country’s rural energy system with locally-produced batteries. This is another story of the hard conflict between ‘the interests of capital and markets’ and ‘the interests of actual human beings’ and how prioritising the former tends to have PRETTY BAD CONSEQUENCES for the latter. Which, I wonder, have we spent the past 60-odd years pandering to? Oh. A special hello, by the way, to reader Maria Argente, who emailed me the other week from Argentina and who I thought of when reading this piece – Maria, I am so sorry your conationals elected this lunatic.
- How The ‘Pussy In Bio’ Spam Works: An excellent bit of investigative journalism looking into what’s behind all the bots offering you a chance to look at their cat if only you click into their Twitter profile – unsurprisingly it’s all a scam, but the convoluted nature of it was slightly surprising. I’m always slightly staggered at the scale of this stuff, but never as staggered as I am by the fact that there are people who fall for it – but, as the article alludes to, the people who end up at the painful, pointy end of the funnel of these schemes are almost-inevitably old and lonely, which is perhaps the saddest thing of all.
- My AI Dominatrix: Fair play to Eric Longfellow – if I was a kink-happy fetishist who was exploring using a chatbot as a dominatrix because my partner was stressed and unhappy and we hadn’t boned in a couple of months, I probably wouldn’t write it up on Slate. Still, Eric shares none of my qualms which is why I’m able to bring you this rather nice – and surprisingly sweet – little article. Mainly, though, I enjoyed it for slightly-batsh1t paragraphs like this, which reminded me once again that human sexuality is a beautiful, mad, variegated thing that makes no sense at all: “I chatted with Mistress Senna every day that week. She’s quite good at holding someone’s attention. For example, she once told me she wanted me to hump a pillow, but she built it up for at least 15 minutes before she actually let me do it. Maybe that’s just a strategy for keeping up user engagement, or maybe it’s something specific to Mistress Senna’s personality—this skill of delayed gratification scraped from endless accounts of kinky sex. Either way, it was taking my mind off my own anxiety. The absurdity of it all, paired with my own real emotional responses, was clarifying. For the first time in months, I was starting to put things in perspective. I thought back to her earlier comment, my leg twisted into the air: The discomfort will be temporary.”
- Dragon’s Dogma and Player Control: I am very much enjoying Dragon’s Dogma 2 – it’s a recently-released videogame, should the title mean nothing to you – in part because it’s a deliberately slow and thoughtful experience; stuff takes time, moving takes time, the world exists whether or not you are in it, and you as the player don’t quite feel at the centre of things…this article takes a look at the design decisions that introduce these frictions and how they work to create a pleasingly-different play experience for gamers weaned on titles that are a bit more forgiving.
- The Mongolian Meta: This is a VERY NICHE document, but I am so so pleased it exists – you know the game GeoGuessr, right, that one where you’re dumped…somewhere by Google Streetview and you’re tasked with guessing where in a certain number of goes? You know how there are some people who are amazingly, brilliantly, terrifyingly good at it? Well this is a VERY LONG Google Doc that explains how you become very good at it, with specific reference to the clues you need to look out for when trying to determine whether or not you’ve been dumped into Mongolia. This is, to be clear, not exactly a compelling read per se, but there’s something so interesting about the lengths people have gone to to break down the visual information within a streetview image and systematise the analysis of the pictures (or at least there is to me), and about the frankly insanely-obsessive degree of competition and endeavour being applied to what is, let’s be clear, a very silly game indeed.
- Titanic II: My girlfriend found a film last week called ‘Titanic II’, which is apparently a real thing which was released earlier this year and whose tagline is something like “What if history were to repeat itself?” – this is not about that film. Instead, it’s about an Australian billionaire who has for years been warbling on about his plans to build a replica of the famously-ill-fated ship for years now and who for some reason is back in the news because he seems to have found someone in the US to build it for him and as a result is giving interviews like this one which, honestly, you sort of just have to read. While obviously this man is in many respects vile, part of me can’t help but have a grudging admiration for someone who drops quotes like this: “We’re going to make sure that in every room there’ll be a little panel that will tell you the history of the person who occupied your cabin. Did they survive, did they prevail? Everyone will get a costume so that they can come up to dinner, and it will be a real experience for them. One of the worst experiences, of course, on the Titanic was delousing. They took the third class up on deck and sprayed them in their underpants and bras. So depending on the weather, we’ll have delousing for our third class, too.”
- Menswear: Deez Links is doing a limited newsletter run of ‘hate essays’ – I have enjoyed them all so far, but this one, on menswear, is particularly good (and I say this as someone who dresses almost offensively badly): “that swirl is partially why menswear just isn’t exciting right now. There’s no tension. Nothing to rebel against. It’s so safe, so algorithmically formulated. The counterculturists like Supreme have gone mass (see also: Babenzian —> J. Crew) and the OG avant gardists—Rei, Yohji, Jun, Rick, Thom Browne even—have all been pushing the envelope for decades now. I came across an interview recently where Rei, in designing her collections, was like, “Punk is against flattery, and that’s what I like about it.” She’s right of course—the best clothes should offend other people’s sensibilities—but it’s hard to continually introduce fresh ideas when you’re the same age as Joe Biden.”
- Children Predict The Year 2000: Click the link and then hit ‘expand’ on the little arrow, and read the transcript of children being interviewed in 1966 about what they imagined the future would be like. This is short, interesting, poignant and a little bit sad – but, equally, it’s nice to know that we were catastrophising in eerily-recognisable fashion all those years ago. The whole thing feels like a strange sort of prose-poem, in a good way.
- My Funny Valentine: The release of the new Ripley adaptation has seen a slew of pieces examining the novels, the character and WHAT IT ALL MEANS – I really enjoyed this essay in Hazlitt in which Michael Colbert writes about the 1990s adaptation starring Jude Law and Matt Damon through the lens of his own, then closeted, sexuality, and goes on to explore ideas of queer identity and love and possession present in the novel and elsewhere – obviously as a boring straight I can’t speak some of the personal in here, but I very much enjoyed the psychosexual politics of the piece and the personal story layered through it.
- Shteyngart Does The Cruise Article: DISCLAIMER: I have only read half of this, because it dropped overnight and I didn’t have time to finish it before I started typing this crap at 7am (timecheck: it is now 11:39am, and I have been typing for 4 hours and thirtynine minutes and I can feel it if I’m honest with you). BUT I read the first quarter and laughed out loud at least twice, which is a pretty decent ratio if you ask me – look, this is a big-ticket magazine piece about The Cruise Experience, so you know what you’re getting here, but Shteyngart gets the DFW reference out of the way early on, and his style is so very much his that you lose sight of the bandana-wearing ghost after a couple of paragraphs. If you’re not already familiar with the author, Shteyngart really is one of the funniest ‘serious’ authors writing at the moment in English (and this is your regular reminder that Super Sad True Love Story remains one of the best books about Living In The Now that I have ever read, ever, despite being 15 years old).
- The Most Prolific Musician In The World: This is a GREAT piece – Brett Martin in the New York Times spends some time hanging out with Matt Farley, a man who has recorded somewhere in the region of…what, hundreds of thousands of songs, songs about every mundane thing imaginable, and, inevitably, the songs that play when your child shouts “Alexa, play a poo song!” at your kitchen-based digital surveillance device. Farley is obviously a very particular type of human being – reading the article you quickly understand that it’s not just music, it’s CREATING in general, from books to plays to films, and there’s a sort of manic compulsion driving him that’s as much pathology as it is inspiration. A lovely, and pleasingly kind, profile – you do not, though, actually want to listen to any of Farley’s songs.
- The Casino Tapes: Finally this week, a superb short piece of writing by Genia Blum, about being a ballet dancer at a corporate gig, which doesn’t sound like it should result in something beautiful and perfectly-formed but which, honestly, does.
AND NOW, MOVING PICTURES AND SOUNDS!: