Sincere thanks to all of you who emailed me last week to, er, check that I wasn’t about to do some sort of internet-induced harm to myself – don’t worry! If I do *that* I’ll put the livestream link in that week’s Curios so that ‘MY COMMUNITY’ can join in the fun!
(seriously though, it’s all authorial pose! I love the web! It’s not like some sort of appalling addiction or dependency or something, or a gaping void in my life that I’m filling with information in the absence of anything else that works!)
I hope you’ve all had good weeks, in any case, and that you have fun things planned for the weekend – I had to spend more of it than I would ordinarily have liked staring into the abyss of AI-generated bongo (no, really, I did! For an actual job!) and so I am going to go out now and gaze at the waters of the Thames in the potentially-vain hope that they will somehow cause me to forget all the terrible, dead-eyed ‘sexy’ pictures.
I am still Matt, this is still Web Curios, and you will almost certainly want to thank me for the Now! playlist.
By Amie Dicke
- Orbit: I fcuking love the BBC. Really, I do, I can’t help it – I have the sort of generational attachment to the Corporation that can only come from having been parented by it for significant proportions of my life (this is, of course, MASSIVELY UNFAIR on my poor single mother who obviously had to, you know, go to work and pay the mortgage and probably, if I’m honest, get away from me, but, well, she’s dead and won’t see this, so). It’s educated me and entertained me and occasionally paid me money, and it’s one of the few objectively ‘good’ things we are renowned for internationally as a country (the royal family is not an objectively ‘good’ thing, nor are the Beckhams, nor is OnlyFans, and I think we’re all agreed on the at-best ambiguous nature of the whole ‘centuries of colonialism’ project), and it also contains people who experiment and play around and make occasional things like this, and well, it is worth a tenner a month is all I’m saying. ANYWAY, Orbit is a new experimental webtoy thing, made by the BBC and designed to offer a new way of discovering music – each day you can go to the website and are presented with a circular interface (a bit like an orbital diagram, DO YOU SEE?) which lets you hear snippets of different songs (the longer you click on each ‘planet’ in the orbit, the more of the song you hear. If you like a song fragment, the software suggests a selection of others which naturally flow from it in a ‘soundalike’ fashion, letting you pick up to 5 new songs each day to add to your personal playlist – you can then add tracks to your playlists on various streaming services. Effectively this is a daily guided music exploration tool, a sort of ‘follow your ears’-type affair, but it has SUCH a nice interface and I think it’s a really nice way to spend 5 minutes each morning discovering a few new songs. Would Sky have made this? Would they fcuk.
- Verse: One of the many, many cultural predictions that I have confidently made and which have never come anywhere near to becoming reality – or at least not in any meaningful sense – was about…ooh, 7 years ago, ish, when I firmly believed that the ubiquity of the ‘stories’ format across social would lead to a resurgence in self-publishing and zines and things of that ilk (I maintain that this is *nearly*, almost true-ish, if you squint, but also concede that I might be talking bunkum). Anyway, maybe this time’s the charm – Verse is an interesting-looking new app (iOS-only, annoyingly, which means my impressions are entirely based on, er, reading a few articles about it and looking at the app store listing) which basically lets you spin up little webpages which you can then share via socials. It’s slightly hard to describe, but I can see a clear throughline between the aesthetic here and that which was all over Stories a few years back; basically you have a blank canvas onto which you can dump text, images, short video/gifs and the like, link out to other places (ie Reels/TikToks, etc), and which in general have a sort of ‘scrapbook/fanpage’ vibe. It feels VERY teenage girl in that sense, but also quite…fun, and lightly-creative, and the sort of thing that you could probably do quite a lot of fun stuff with with a bit of thought and effort. Aside from anything else, the thought of pulling one of these together on my phone makes me feel genuinely ill – HOW DO THE CHILDREN DO THIS STUFF WITHOUT A MOUSE AND KEYBOARD? So old, so desiccated, so left-behind.
- Gentype:Would you like your very own bespoke font, made by AI (courtesy of Google)? OF COURSE YOU WOULD! This uses…oh, fcuk’s sake, I honestly cannot keep up with the various different model names all these fcukers are deploying, nor indeed do I care. It uses SOME FCUKING MAGIC GOOGLE TECH (I checked in the end – it’s Imagen2, should you care) to create an entire A-Z font in whatever style you prompt it to – which, obviously, is guardrailed to high heaven and won’t let you create anything TOO contentious, but, as ever, you can get around this to an extent (‘viscera’ it doesn’t like, but I found that ‘tendons’ produced a similarly (and unpleasantly) meaty result, for example). This is obviously just a silly toy, but it’s impressive how well the outputs work, and you can download your alphabet should you want to preserve your masterworks forever, or even just type out a message and get an image of that should you wish to, I don’t know, create a font out of dead flowers and send “LOOK WHAT YOU HAVE DONE TO OUR LOVE?” to your significant other (don’t do this, though).
- AI Or Not AI?: Another of the NYT’s occasional ‘so, can you tell what’s real any more?’ quizzes – the link here takes you to an archive version so the ‘quiz’ element doesn’t actually work very well (if you’ve not yet been paywallblocked then just go to the original url and play as normal), but you can still see the videos here and make your own guesses as to which are real and which are spun up by Kling etc. I have to confess, I got one of these wrong (don’t worry, the beatings were severe) – which, for someone who spends as much time as I do staring at this sh1t, was frankly embarrassing but also testament to how much better this stuff has gotten (but, to be clear, only for VERY short clips – anything longer than about 3-5s and the illusion is shattered). See how you get on – you will, I think, be surprised.
- DebunkBot: Ooh, this is interesting. A paper was published by MIT this week seeming to demonstrate that prolonged interaction with a specially-trained LLM was able to effectively help in debunking conspiracy theories, or at least partially doing so, amongst people who had previously held strong opinions about, say, COVID, or the US election in 2020. You can read the paper here if you’re interested, or you can click the main link, roll up your sleeves and have a DEBATE about, I don’t know, chemtrails. As a general rule I like to think of Web Curios readers (lol, of course I don’t, I get really upset if I ever conceive of your existence, I do this for ME) as rational, coherent, *intelligent* people who wouldn’t ever engage in MAD BELIEFS – equally, though, it’s been a long few years and we’ve all spent a lot of time online by now, so I suppose it’s entirely possible that at least a few of you believe that Hillary Clinton and Tony Blair and George Soros all really do run an international adrenochrome-farming operation involving a complicated and deeply-Satanic process of child gland harvesting; if that’s YOU, why not tell DebunkBot all about it and see if it can’t wash some of the madness away? I am curious, though, as to how this gets past the (to me obvious) objection of ‘well, of course, you’ve programmed the machine to think this stuff, it’s part of the same establishment conspiracy and you’re probably all satanic paedos too’ – anyway, let me know how you get on.
- The Paralympics Australia Virtual Stadium: On the one hand, I have no desire to mock the Olympics or the Paralympics, which by all accounts were both amazing and wonderful and inspiring and all that jazz; on the other, SOMEONE DONE A METAVERSE! Ok, they don’t actually describe it as such, but you can bet that when this was initially pitched back in…ooh, thinking about timelines and the like, let’s say late-2021, that the ‘M’ word was 100% on one of the slides. Anyway, this is an IMMERSIVE DIGITAL STADIUM EXPERIENCE launched to enable Aussies to get behind their Paralympians by visiting a virtual version of an Olympic stadium – you know the deal by now, WASD to move your avatar around, space to jump…sadly there’s no button to make the experience ‘good’. You can…what can you do? You can walk around the arena, approaching various BIG SCREENS on which you can see low-res clips of that day’s Aussie highlights (which are obviously not updating anymore because, well, it’s over), and there’s a merch shop (OBVS) and you can also show your support for the Paralympian team and Aussie sport in general by BUYING A VIRTUAL SEAT in the stadium for $25 actual cashmoney (or a virtual corporate box, if you were feeling particularly flush). The thing is, I obviously saw this and LOLed, but then explored further and it seems that they’ve sold 31,000 of the fcukking things, raising $800k, and I realised I know nothing. Australians reading this – does that sound plausible? Do you know anyone who’s bought a ‘virtual seat’ to the ‘virtual paralympics’? Because I have to say I find the idea of 31k people rushing to spend 25 quid on this astonishing in the extreme, but they couldn’t be lying about this, could they?
- Dynamicland: Look, I don’t really understand what this is AT ALL – the FAQ reads “an independent nonprofit research lab, whose mission is to enable universal literacy in a humane dynamic medium. This involves inventing a humane form of computing, and developing educational and community-based institutions in which a culture can grow”, but, I’ll be honest, these are words whose meanings in isolation I comprehend but which when arranged like that baffle me entirely. Still, that’s not the point! The point is that I absolutely adore their website, which is based around a photo of some shelves and which, honestly, is SO CHARMING and I now think that everyone should make websites like this – fcuk complicated CSS, fcuk Java, fcuk all this complex stuff, let’s just chuck up a photo with some hyperlinks and be done with it (obviously I don’t mean that, but there’s something really pleasing about the simplicity and the design here).
- The Salad Fingers Shop: I generally try not to link to stuff that’s ‘just’ for sale here (a rule I break I think three times in this edition – I am nothing if not entirely inconsistent and largely-hypocritical!), but I will make an extra-special exception for the newly-launched Salad Fingers online store. If you don’t know what Salad Fingers is then what the fcuk are you doing here, frankly (but also, educate yourself)? If you do, then, well, you might find some of the merch available here of interest – the latex costume mask is a particularly-horrifying (and to my mind decent value) addition to your Hallowe’en/erotic (delete per preference) arsenal.
- 575 Life: Many years ago, when I had probably the least-healthy job of my life, I was in charge of doing the morning meeting notes at a certain PR agency; because I am a pr1ck, I took to starting each one with a TOPICAL HAIKU (personal favourite: “Welcome Suri Cruise! / Real child or creepy changeling? / Only time will tell”) – as such, the form has a special and permanent place in the wizened cabinet of horrors that is my ‘heart’. 575 Life is a project that posts haiku – I don’t know who by, I don’t know how often, and I don’t know why, but I rather like them, particularly “four letter words can / fill to the brim or empty / your entire soul.” You can sign up for occasional haiku updates, should you so desire – why don’t you subscribe? / It probably won’t hurt you / (though I can’t promise).
- Cellar Door (Redux): So a few weeks back I posted a link to Cellar Door, a project seeking to find the BEST (or at least most-popular) word in the English language, and made some sort of typically-churlish complaint about how I was annoyed at the fact that there wasn’t something trying to find the WORST word – and then obviously one of the nice people behind the website (Aris Catsambas, in fact) got in touch to say that, actually, that does exist, and kindly sent me the link. So now we can learn that, per the ongoing experiment, the least-popular word in the English language at the time of writing is…GRAPEFRUIT, apparently, which seems to me a bit weird but at least it’s not fcuking ‘moist’. Although the fact that ‘serb’ is second makes me wonder what metric’s being used here and whether, er, there isn’t some sort of weird ethnonationalist conflict being played out via the medium of wordpreference. This list shows you the bottom-100 words as ranked by strangers on the internet; personally-speaking I’d pick ‘wert’ (also, I like to imagine ‘robberies’ is on there as a result of persistent downvoting by people with minor speech impediments).
- Obsolete Sony: TOO MANY FCUKING NEWSLETTERS. Still, this is the sort of thing that you will either immediately pass over because, well, life’s too short, or which alternatively you will spend the rest of the afternoon reading the back-issues of. Would YOU like a regular newsletter detailing all the different product ranges and models that Sony has launched and then subsequently shuttered over the years (example subject line: “QUALIA: Sony’s Most Expensive Luxury Series 2003-6)? For 99% of you the answer to that question will inevitably ‘no Matt, fcuk off, why would I want that?’ – but you, the other 1%, the MAD ONES, the ones mad to live and love and reminisce about previous generations of electronic hardware, this is for YOU!
- Olana Light: Via Blort, Olana Light is a UK artist who…oh, look, here’s her explanatory blurb. “My practice moves between wearable sculpture performance, installation, photography, and the moving image, and reflects the multiplicities of identity and a never-ending pursuit of belonging that are close to my own heart. Exploring notions of ‘self’, and its connection with the body and nature, my work offers new perceptions by challenging audiences to accept the absurdity of the ‘other’, to question their beliefs, and to interrogate their own sense of belonging. I seek to create a dialogue for change: about nature and our relationship with it; about who we are; and about why art is and should be for everyone to access.” Got that? Good. Now click the link, click ‘Projects’, and MARVEL at the beauty of the costumes and the design and the craft and the work here. This is SO beautiful, and I now want to dress up as a birch tree.
- Elle’s Home: A lovely little personal homepage website – hi Elle, whoever you may be! – which is packed full of cute features but which made me fall in love with it because she’s implemented Matt Webb’s ‘cursor chat’ tech which means that everyone currently on the website can see each others’ cursor and, if they so choose, chat to each other while they browse, and I love this SO much (multiplayer websites! A trend! PUT IT IN YOUR FCUKING 2025 ‘DECKS’ YOU CNUTS!) and I think you will too.
By Pale Flare
- Cada: Ok, the second of this week’s ‘links which are basically just taking you to a shop, for which apologies again but, well, it’s practically Christmas (SORRY) and you will probably need to think about presents soon, so actually I am doing you a favour what the FCUK are you complaining about?’, this is something which may well be OLD NEWS to people who are either parents or modelmaking enthusiasts but which to me was an exciting new world of modular bricks. Cada appears to be a Chinese(?) model-making system, almost-but-not-totally-unlike LEGO – it seems a *bit* more complex, but the real draw (for me, at least) was the odd licensing deals they appear to have made with real-world companies like, er, Shell. Want to have a complicated, multi-part assembly model of, er, an actual Shell forecourt? GREAT! There are also OFFICIALLY LICENSED models of all sorts of car brands and the like, should that be more your thing, but I personally couldn’t get over the image of diminutive plutocrat Wael Sawan having the licensed Shell motorway services shop on his desk and playing with it contemplatively while his extractive minions set the world on fire.
- Things In…: Ordinarily I don’t believe AI travel planning services are in any way a valuable thing – all they seem to do is point you towards either the most obvious and popular destinations in any given city, or alternatively ones that don’t in fact exist – but this seems to work…marginally better. No idea how it’s working under the hood, but it feels reasonable to assume that there’s some vague LLM-related plumbing under the hood. Anyway, tell it where you are going and it will think a bit and then spit out a bunch of recommendations around visitor attractions, food, places to stay and general tips. If you try it for a big city it will be very generic (although I thought the restaurant recommendations for London were better than you usually get with these things), but if you try it on somewhere more obscure it’s surprisingly not-terrible. I *think* it’s pulling stuff based on a combination of LLM suggestions and Google reviews(?), but I would be genuinely fascinated to know quite how it’s all put together.
- Great Ball Contraptions: It is a source of no little shame to me as a man of 44 years that I started sniggering to myself as I typed those words, but, well, I also think it’s important to be honest about one’s failings. Have you ever wanted to “participate in the fun of Great Ball Contraptions”? Would, er, you like some more detail as to exactly what ‘great ball contraptions’ are? Here: “A great ball contraption (GBC) is a machine which receives soccer balls or basketballs from one module and passes them to another module, rather like a bucket brigade. Modules built according to the GBC standard can be assembled into a collaborative display without pre-planning or modification. The GBC standard is minimal, permitting the builder great flexibility in designing the mechanism by which balls are moved from the input to the output. The otherwise pointless handling of balls, and the myriad ways this is accomplished, gives great ball contraptions the impression of a Rube Goldberg machine.” So basically a GBC (if I only type the initials I don’t start giggling, turns out) is a machine designed to move stuff, pointlessly, often made of LEGO or similar building system. This is sadly light on videos, but there are LOTS of instructions should you want to spend this weekend monopolising your kids’ LEGO sets for a project that will not interest them in the slightest while you ignore them and their lonely tears.
- 50Watts: I am pretty sure that I featured this YEARS ago when it was on a different url, but I was sent it this week by reader Hyunsuh Kim (thankyou Hyunsuh!) and it’s lovely. It’s “a growing archive of weird and wonderful visual ephemera from around the world”, and there is SO MUCH rich visual inspiration in here, with hundreds of posts themed around a particular artist or illustrator or theme or style, and if you’re someone who works in publishing or design (or, er, design for publishing!) then I think this will be hugely up your street.
- TuneShine: Our final (promise) nakedly-commercial link of the week is this – an honestly very, very cool piece of product design which I can imagine having a certain number of middle-aged dads amongst you reaching for their wallets (do YOU live in Walthamstow? Do YOU call it ‘the Village’? Yes, YOU!). TuneShine is basically an LED art display – handmade (they know their audience, these people) in LOVELY WOOD with an LED-enabled front panel, this basically displays a lo-fi, lo-res version of the album art for whatever’s playing on your streaming service of choice (Spotify, Apple Music, etc) – effectively a sort of domestic artwork/what’s on display, a bit like a digital version of those stands on which you put the cover of whatever vinyl you’re currently playing. Very much the sort of thing that you might have found in a dot com office in Shoreditch c.2010-13 (RIP TECH CITY) which would inevitably been ruined when some bright spark decided to do mephedrone off it after coming back to the office for afters following Silicon Drinkabout. Anyway, one for your Christmas lists – it’s not cheap at 200 dollars, but it’s *very* pretty.
- The Railway Movie Database: Ok, so I had this on the list this week but I confess to not having properly dug into it just now (yes, this is a glimpse behind the authorial curtain – you SEE the sort of preparation that goes into every edition of Web Curios? This is why I could never ask for money for this sh1t), and I just checked the ‘About’ section and got unexpectedly incredibly emo about it (it’s 9:04am, this is normally when the fatiguetears start to come in waves). It is clear that this person REALLY CARES about trains and railways, and as a result of that intense passion has decided to create, and maintain, this website which contains an absolutely INCREDIBLE amount of information about the train stations of the UK and Ireland and which, if any, films and television series they have featured in. If I were to quibble I would like the ability to be able to browse by station rather than by film – THIS IS NOT A COMPLAINT, just a statement of preference! – but, honestly, I am slightly in awe of the amount of effort that has obviously gone into this, and the fact that it’s still being maintained in 2024. The person behind this is one Jonathan Horswell, who I think is doing God’s work in some small way.
- Things My Girlfriend and I Have Argued About: Do you remember the days before every fcuker in the world had a fcuking podcast? God, they were good days, weren’t they? Anyway, in that strange, halcyon period, rather than recording baggy, poorly-edited and unfunny conversations between friends that noone else never, ever needed to hear, people instead committed their not-particularly-funny musings to…er…baggy, poorly-edited webpages instead! I have a strange feeling that this website might have been VERY FAMOUS in the early days of the mass internet – one of the original viral sensations – but I have only a fleeting recollection and, honestly, I am not going to Google this because there’s every likelihood that I would fall into a terrible researchhole and never emerge. Still, this website from…what, the late-90s/early-00s? Anyway, it collects things that the author (apparently English) and his girlfriend (German) have argued about – I appreciate that what I am about to type here is a bit glass houses, for which apologies, but the writing isn’t very funny and the observations are kind-of banal…I find it interesting, though, because a) it’s an archive of a DIFFERENT TIME (I do not think the author would have enjoyed going viral with a lot of these observations in 2024, put it that way); b) it’s an interesting artefact of THE PAST WEB; and c) it’s literally EXACTLY the contents of a significant number of podcasts, many of them hosted by couples, which proves that there is nothing new under the sun and that this sh1t has honestly never been that funny so can we please stop now? Thanks.
- A Single Div: An OLD design project by Lynn Fisher, demonstrating a bunch of pretty things you can do with a well-crafted bit of CSS. There’s some lovely webdesign stuff in here which might prove pleasingly-inspirational; alternatively you can just scroll and stare, slack-jawed, at PRETTY THINGS ON YOUR SCREEN, which is very much my approach to it.
- Nights On Earth: Ooh, I like this rather a lot. As the nights are drawing in – I woke up this morning to write Curios and it was PRACTICALLY DARK, which, honestly, I wasn’t wholly psychologically ready for – it seems appropriate to share something you can do with the night skies (although in England in winter those night skies tend to be accompanied by heavy cloud cover so, er, perhaps this is a pointless endeavour). Click the link, tell the site where you are, and it will present you with a *really* nicely-designed monthly calendar view, where on each day you’ll see a small graphical guide to what you might expect to see in the nigh sky that night (presuming, you know, there are no clouds, and you live in the middle of nowhere rather than in central London where it never actually gets dark properly) – so for example I can tell you that there’s an increased chance of aurora events tonight (not that we’ll see them, but), or that there’s a full moon AND supermoon next Wednesday, that sort of thing. Aside from anything else, the design/webwork on this is really very pleasing indeed – also, though, it makes me quite upset that I haven’t seen any shooting stars this year, might have to rectify that.
- Sunlitt: Seeing as we’re doing ‘celestial stuff’ (SEAMLESS, I tell you), here’s an app which will show you the position of the sun (and the relative shadows cast by buildings anywhere in the world) at any point in the day – this is SO PRETTY and so nicely-designed (or at least it *looks* pretty and nicely-designed, it’s another iOS-only app, the fcukers), and while I can’t immediately think why you would need this information with you wherever you go I like to imagine that at least one of you will be able to find some sort of use for it (planning photoshoots? Beach trips? Walks when on holiday? Yeah, ok, that sort of thing).
- Presidential Ham: As we wait with bated breath for the fcuking Americans to finally do their fcuking democracy and stop sucking up all of the rest of the world’s news oxygen, why not pass the time with this gallery featuring paintings of every single US President, each accessorised with, er, a ham. George Washington, holding a ham! Grover Cleveland, holding a ham! Barack Obama, holding a ham! WHO WILL HOLD THE HAM NEXT? There’s a ‘Joe Biden holding a ham…coming soon!’ placeholder at the bottom of the page, suggesting both the site owner’s political leanings and that they probably need to update this – but, honestly, here’s hoping that come December there’s a new portrait of a ham-laden Harris gracing the site.
- Cleanerbot Rescue: A reader (Andres Varela) writes!: “I built a thing. It’s an old school text adventure, only it’s voice controlled. I built it to learn some stuff and be a proof-of-life to folks in my network. Turns out they like the writing, and keep playing, so I thought I’d get cocky and see if you’d like it too…it applies AI to divine user intent rather than force them to use overly specific command phrases from the late 1900’s. “Have a look around and tell me what you can see.” vs “search”. There’s a game save function that not enough people cotton on to, which I think is nifty because it doesn’t require players to register.” THANKYOU ANDRES! I initially struggled a bit with this because I was focused on using old-school text adventure commands, but once you lean into the LLMiness (IT IS A WORD) of it then you will get the hang quite quickly – it’s fun, especially if you’re old enough to remember this stuff the first time around, and while, if I’m honest, the voice commands are just a gimmick and it would be quicker to play by typing, there’s still something slightly magical about the fact I can speak to my computer and play (admittedly rudimentary) games by shouting at it.
- Ducky Fog: This week’s final frivolity is this tiny game in which you have to move the duck and rotate the screen to reach each level’s goal. Don’t worry, it will all make PERFECT SENSE as soon as you click the link, I promise (oh, and if you want you can mess with the code too, but, er, I don’t really understand code and so I can’t comment on that bit).
By Ludwig Favre
THE CIRCUS OF TUMBLRS!
- One Terabyte of Kilobyte Age Photo Op: Oh SUCH nostalgia! “This tumblr accompanies One Terabyte of Kilobyte Age, the Geocities research blog by
- Olia Lialina and Dragan Espenschied. Screenshots are automatically generated from a stash of old Geocities home pages, rescued by the Archive Team in 2009. The files are processed from oldest to newest.” All of these are dead now, making this a sort of digital graveyard, but it’s a gorgeous reminder of that brief period when ‘being online’ meant ‘creating a space that you felt expressed who you were and using that space to find and connect with other like-minded people’ rather than ‘creating a funnel through which a variety of different billionaire corporations can slide a variety of pabulum directly down your infogullet’.
THE TROUGH OF (INSTA) FEEDS!
- Jack Davison: Photos by someone called Jack Davison. GOOD photos, should you need an additional reason to click – nice range of subjects and styles, and in general more interesting than your standard insta ‘landscapes and portraits’ feed.
- The Daily Splice: A feed featuring a load of really quite excellent collage-y images and animations – this is some really impressive work, and stylistically very coherent to boot.
- Minimus: This is the Insta account of Minimus, according to the bio ‘The Mouse that made Latin cool!’. Erm, did it? I confess to not having noticed the now-incredibly-hip status of the dead language, but perhaps it’s because I’m not cool and in fact what all The Kids are doing rather than going and getting fcuked-up on meths is, I don’t know, declining verbs and engaging in hot conjugation sessions with their friends. I doubt it, though. Still, there is a LOT to love about this, not least how adorably shonky the little mouse doll that features in all of the images is, like something from a particularly low-budget BBC kids show from the 70s. Minime, rex es inter mures et te saluto!
LONG THINGS THAT ARE LONG!
- Art Project: Our first link this week is honestly FASCINATING – I promise, even if you don’t care about AI, that this is a really interesting exploration about what it means to make art, to think, to explore concepts, all of that chewy stuff. Frank Lantz, whose writing I’ve featured in here before, takes the provocation from the headline of last week’s Chiang piece (‘AI Can’t Make Art’) and runs with it – this article is a series of screenshots of the conversation between Lantz and GPT-4o, in which Lantz attempts to patiently train The Machine into making something that is ‘art’ based on his definition, and as a glimpse into how LLMs ‘think’ – and specifically how they actually very much do not ‘think’ at all – it is SO SO SO GOOD. I can’t pull excerpts because the conversations presented as screencaps, but trust me when I say that it’s one of the best explorations of the possibilities and limits of this sort of technology that I’ve yet seen, as well as being a generally fascinating exploration of the wider concept of ‘what is art?’ and what is (and isn’t) possible with the current generation of models, and how, as per previous Curios, they are excellent at mimicking the ‘shape’ of thought and meaning without in fact being vessels for actual thought and meaning in any sense whatsoever. BONUS, TANGENTIALLY-RELATED LINK: this piece in the Atlantic, by the guy who did that viral Tweet about replacing all the icons on his phone with AI-designed ones featuring Kermit, touches on a few vaguely-similar points (albeit more accidentally) and is a nice, lightweight companion to the BIG THINKING above.
- The New OpenAI Model: Except maybe now it’s all going to change (I don’t think it is, you know)! Maybe the new OpenAI model – o1, released yesterday but only being rolled out slowly to developers at first, so most of us won’t have had a go with it yet – will solve all this stuff (it won’t)! The main link here is to the VERY DRY OpenAI announcement blogpost which contains all the info about how it’s performed against all the usual benchmarks but which, honestly, is a bit dry – it does, though, do a reasonable job of explaining the BIG DIFFERENCE, which is ‘reasoning’. Basically what that means is that the model responds more slowly, but does so in a way that works through various interrogatory layers to your original prompt, effectively building in a lot of the conversational hoops you need to jump through to get decent work out of LLMs under the hood. To get a better understanding of what this looks like in practice we can turn to the ever-wonderful Ethan Mollick, who’s apparently been messing with it for a month pre-release and in this blogpost gives you some practical examples of how it works and what it can do that previous models would have struggled with. The TL;DR here is, seemingly, ‘this will do better at tasks that are ‘think-y’, that require a bit of ‘if this then that’-type reasoning’ – and I have to say, this example really did impress me. Although, per Sam Altman himself, this isn’t AGI, it’s nowhere close, and it becomes less impressive the more time you spend with it and work out where the gaps in its abilities and performance of ‘intelligence’ lie, so, well, there you go.
- An Interview With An Anthropic Person: Sorry for all the AI stuff uptop this week – I appreciate that some of you are, er, not HUGE fans – but this is another genuinely interesting piece; The Verge interviewed Anthropic’s Head of Product about all sorts of things pertaining to AI, with a refreshing focus on, you know, delivering product and ‘what is this for?’ and ‘how do we think about safety?’ rather than the more woo-woo ‘so, is AGI coming and will it kill us?’ bullsh1t that often characterises some of these chats. Ok, so there is a *brief* segue into an imagined future in which we Infinite Jest ourselves into total civilisational collapse, but otherwise this is measured, interesting and practical (oh, and it’;s available as a podcast too for those lazy deviants amongst you who prefer that as a medium. You’re wrong, but I will still pander to you because I am nice and tolerant and I need you to love me).
- The AI and Energy Question: This is REALLY interesting but also, I must confess, so far above my head from a technical point of view that it may as well be some sort of satellite. Anyway, this is the first in a two-part article examining IN DETAIL (I mean it) the question of ‘how much energy is all this AI stuff using anyway?’ – the second part is linked to from the bottom of the first, and without wishing to spoil the ending here, the answer (with a LOT of caveats) appears to be ‘quite possibly a bit less than we might initially have thought’. Which is not to say that the tech is anything other than energy-intensive, but that you might want to retire the ‘so, you boiled a lake to make that busty Garfield, eh cowboy?’ snark for a bit.
- Information Foraging: Ok, this is *quite* geeky/technical/academic, and technically it’s a webdesign/UX-focused piece, but I think there’s a lot of interesting stuff here which could be usefully applied across all sorts of different fields and disciplines and which broadly falls into the category of ‘interesting stuff that those of you who like to call yourselves ‘strategists’ will probably find tangentially-useful and which, even if you don’t, will make you FEEL as though you did’. The central premise is here – as you can see, you can use this sort of stuff for everything from personabuilding to webpage design and LOADS of things inbetween: “Information foraging is the fundamental theory of how people navigate on the web to satisfy an information need. It essentially says that, when users have a certain information goal, they assess the information that they can extract from any candidate source of information relative to the cost involved in extracting that information and choose one or several candidate sources so that they maximize the ratio: “Rate of gain = Information value / Cost associated with obtaining that information.” In other words, if people have a question, they will decide which webpage to go to based on (1) how likely it is that the page will provide an answer to their question, and (2) how long it’s going to take to get the answer if they go to that page.”
- A Trump Profile: I know, I know – YOU HAVE READ ENOUGH! This one though really is a particularly fine example of the genre – Olivia Nuzzi, who has interviewed the fcuker enough times over the past few years to have a degree of access that’s not afforded to every reporter, and with whom it’s obvious Trump feels a degree of comfort, writes beautifully about Where Trump Is Now, in the wake of losing his preferred opponent, and (allegedly) part of his ear, and for the first time in this campaign feeling the tide possibly turning definitively away from him, and it’s a glorious piece, capturing the oddity of its subject and the strangeness of the court around him, the strange Potemkin Village vibe of Mar A Lago and the sense that this might all finally be coming to an end (please God).
- Joining The Petersen School: You may not find it WHOLLY surprising that the educational institution established by Dr Jordan Petersen as a corrective to the woke mind virus infecting campuses across the world has ended up being…a subscription-based service where you watch a bunch of prerecorded videos (seriously, do these people have any other grift?)! Admittedly some of the videos are made my actual academics, and seemingly not all of them are peddling Petersen’s own signature brand of insane, antediluvian gender wars rhetoric, but it’s not entirely clear that $500 a year (in fairness, obviously a fraction of what you would pay for an actual education but, I would argue, it’s not in any meaningful way an actual education, so) buys you anything other than a load of stuff you could probably have found on YouTube anyway, and there’s no suggestion that the ‘accreditations’ one receives at the end of the course have any actual real-world weight, and it does feel a BIT reading this that what Petersen’s done here is to extend his grift of ‘giving really basic advice to the frightened in the form of a self-help book’ to ‘giving really basic advice to the frightened who can’t read in the form of videos’.
- 764 Redux: A while back (23rd February, precision fans!) I featured an unsettling story what I then described as “a group called ‘764’, which, long story short, is comprised of people who get off on making other people do humiliating or harmful things to themselves on camera as a power fetish thing.” This piece is a sort-of followup which focuses on the ringleader of the whole enterprise, a kid from the US called Bradley Cadenhead, now 19 and into year two of an 80 year (!) prison sentence connected to all sorts of unpleasant crimes; look, I can’t pretend that this isn’t just a bit grim, and there’s a certain degree of ‘true crime’ salaciousness to the article that I’m not a huge fan of, but, equally, I always find it fascinating to be reminded of the fact that, actually, this stuff really does happen and it’s not always just myth. Oh, also, a useful additional reminder that it’s not just Telegram where the bad stuff happens (I imagine Discord’s PR team is feeling pretty relieved about how lightly they’ve gotten off, is all).
- Lying For Money On Social Media: In another instance of ‘unintended consequences of poorly-thought-out incentives schemes’, this article examines why there’s a spate of accounts across social platforms – but primarily Twitter – posting stuff that is deliberately wrong or false; the TL;DR here is ‘because there’s nothing more irresistible to a certain type of person than the need to correct someone online’ and that, by posting things that get specific pop culture facts wrong, account owners can leverage this in order to farm engagement, boost virality and coin in some of that sweet, sweet ‘creator income’. But, er, what happens when these viral lies become part of the fabric of the web, and what happens when they get ingested into The Machine and become part of what passes for the informational water table? Is this what the platforms imagined when they started offering cash payouts for popular content? What do you mean ‘they didn’t think about it at all?’ Eh? Oh.
- The AITAverse: Or, to give it its full title, “How “Am I the A$$hole?” ate the internet – this is a look at how the subReddit became so popular, but also at the wider popularity of accounts across all the platforms doing similar blind item social outrage schtick, and why that popularity might be peaking about now. My theory, broadly corroborated by the piece, is that we’re at a stage of online where obviously clowning on actual people feels…bad, and not really ok, and like it might in fact be harmful, but the AITA format is anonymous and possibly made-up anyway, so you can absolutely go to fcuking town on the people involved because they’re unknown, unnamed and possibly fictional anyway. Is this healthy? It doesn’t feel healthy.
- Blog Monetisation:Via Andy, this is an excellent-if-depressing bit of interactive storytelling (I think we can probably kill ‘scrollytelling’ after all these years, can’t we?) about what it takes to make money out of publishing on the internet in 2024. Publishers will, I think, relate to this.
- College Football: Things I don’t understand about America, part x of y – the weird obsession with university sports. Like, don’t you have ENOUGH sport? Why do you need to care about the college kids? Why is your personal identity still stitched to your alma mater despite you being a comfortable two decades from graduation, has nothing else exciting happened to you since then? (NB – these questions are rhetorical, I don’t really care all that much and I appreciate the actual answer is ‘because it feeds into the NBA and it’s a huge part of rural culture, you awful anglo snob’). Anyway, that’s by way of unnecessary preamble to my saying that I don’t understand the first thing about what this piece is wanging on about, but, equally, that it is SUCH a glorious piece of webdesign and it really is worth scrolling through, even as your eyes glaze at the actual words.
- Tennis, Tech, and Gambling: ANOTHER rather nice bit of pagedesign here, this time about tennis and, specifically, the ATP’s arrival in Saudi as part of MBS’ continued determination to make the region about more than just oil, heat and punitive homophobia. It’s focus is on the various technical advancements over the past few years, and how they’ve transformed the sport, and all the interesting and exciting ways in which the data thrown up by the tech can be used to enhance the TRUE meaning of sport, to whit ‘gambling’. A slightly-poignant excerpt: “During the match I observed from the Nest, Hawk-Eye would collect countless data points, much of which was being transmitted live not just to the chair umpire officiating the match but to business partners of the ATP as well — the most lucrative of which are, recently, sports betting companies. Everything that was happening on-court would be sent through an algorithm that would process that information to create more accurate betting odds that could be distributed to the world’s gamblers. This was news to the Hawk-Eye boys. Andrew Birse, a technical project manager, gave me a puzzled look and then got a little defensive: “We mostly deal with on-site capture.” Another operator, Juan Martinez, followed up: “We don’t know what anyone does with it.” I felt bad. They’d had no idea.”
- Uber in Mexico: The story here is pretty much exactly ‘man, I would not want to be an Uber driver in certain parts of Mexico’, but, well, I really wouldn’t want to be an Uber driver in certain parts of Mexico.
- Where Will Games Be In 25 Years: 99% of this is ‘people who work in the games industry offering interesting if (mostly) sober predictions about the sort of technical, ludic or business shifts the industry will see over the coming two-and-a-half decades’, which is interesting but a bit inside baseball. The remaining 1%, though, is the best response I have EVER read to one of these round robin interview questions and you should all read it immediately and then wish that you were as talented as Hannah Nicklin, who has honestly invented a new genre of microfiction-disguised-as-industry-talking-head-comment here.
- How To Build A Walking Table: This is a set of instructions on how to build an actual, honest-to-goodness remote controlled walking table – it will be out of reach for all but the most dedicated and technical of you (and even those people will need access to woodworking tools and a bandsaw), but, trust me, click the link because you will be charmed and you will wish that you had paid more attention in Design Technology all those many years ago.
- Celebrity Number Six Is Found: Are you aware of the internet mystery that’s been surrounding the mythical ‘Celebrity Number Six’ for a year or so? No, of course you’re not, you’re fully-rounded people with actual lives and interests outside of the fcuking internet. BUT it is an interesting story and a sort-of-heartwarming example of collective effort and endeavour (admittedly, fine, entirely pointless effort and endeavour, but still) and the link takes you to Caitlin at Links… explaining what it was all about and why it was interesting in typically-excellent fashion.
- Probability Puzzles: Oh I loved this interview! It’s about the sort of stuff that my brain mostly slides off like fried eggs off teflon (I can just about get my head around Monty Hall, but only if I REALLY think about it, to the point that you see my brow furrow and smoke start coming out of my ears) – probabilities, basically – and about a LOVELY-sounding bloke who’s become moderately-Twitter-famous as a result of posting puzzles. Here’s the intro – I promise, even if your brain, like mine, resists this sort of stuff as a matter of course, this does an excellent job of explaining why these puzzles function as they do, and showing you how to think a bit more smartly about probability overall: “In late January, Daniel Litt posed an innocent probability puzzle on the social media platform X (formerly known as Twitter) — and set a corner of the Twitterverse on fire. Imagine, he wrote, that you have an urn filled with 100 balls, some red and some green. You can’t see inside; all you know is that someone determined the number of red balls by picking a number between zero and 100 from a hat. You reach into the urn and pull out a ball. It’s red. If you now pull out a second ball, is it more likely to be red or green (or are the two colors equally likely)? Of the tens of thousands of people who voted on an answer to Litt’s problem, only about 22% chose correctly.” Honestly, FASCINATING.
- Literary Bratdom: I know, I know. But leave aside the zeitgeist-baiting title (which in this instance is in fact entirely warranted) and this is really good triptych of reviews of three novels published this year, all of which have broadly been marketed as ‘brat lit’ (and one of which even uses the WORD OF THE SUMMER as its title). I’ve only read one of these – the Gabriel Smith one – but I agreed SO HARD with the essay’s assessment of it, and enjoyed the writing and analysis overall, so figured I would include it; even if you’ve not read the books, it’s a really interesting look at YOUNG WRITERS IN 2024, and the whole ‘young writer canon’, and the phenomenon of ‘Brat’…and yes, I know, IKNOW, but I think this is a very good reading of it, cf lines like “trolls were bitter and alienated and politically toxic. Brats are hot, fun, and apolitical; they’ve been feminized and miniaturized and upgraded to a more consumer-friendly model. When they joke, they are not trying to infuriate anyone. They’re selling themselves as aspirationally edgy and unique. They’re flattering an audience that would prefer not to know when it is being pandered to, that would rather believe that it is being tested or confronted. They’re the mash-up of trolldom and capitalism, hoping to elicit a softened form of outrage, an exasperation mingled with admiration and longing. They would like you to believe that their indecision reflects a particular attunement to ambiguity and nuance. But in truth they just won’t know where they stand until they’ve figured out where you do.”
- The NYT Prince Documentary Story: If you haven’t yet read this, you really must – it’s brilliant, about Prince and Who He Was (insofar as it’s possible to meaningfully answer that question), and who we think artists are, and legacy and memory and fame and ‘brand’ and heroes and the concept of genius, and all sorts of other things besides. The central story here is ostensibly about an epic documentary about the artist’s life which is now being contested by Prince’s estate and which may never see the light of day as a result, but it’s FAR more interesting than that.
- Group Chats About Group Chats: A selection of writers write about the oh-so-modern phenomenon that is the Group Chat – you could make, I think, a convincing argument for it being the defining social forum of the decade so far, if you wanted to be a pretentious cnut – and it is SUPERB. I would read a whole edition of the LRB devoted just to this specific topic, fwiw. “This used to be called hanging out. But the difference between the hang and the group chat, aside from plasma, is spontaneity. The hang is now ever-present and constant; it doesn’t require plans. Yet while all the other advents of instant communication—email, breaking news, Twitter, [shudder] Slack—fill me with dread, I can’t wait to open the group chat. It drives my wife crazy: what does the group chat offer that surpasses interfacing with real life? When real life is so slippery, a faceless forum counts for feeling alive.”
- Same River, Same Man: On rereading books at differing stages in your life – specifically, in this case, ‘The Catcher in the Rye’, a book which famously becomes less good the older you are when you read it (other examples of this specific cultural phenomenon – the film ‘Pump Up The Volume’ starring Christian Slater, which is obviously THE BIBLE when you are 15, but which is an astonishingly big red flag should anyone over the age of about 18 cite it as a favourite). Anyway, this is mostly about Catcher, but also touches on various other books and authors – Vonnegut, Updike, etc – and is both a lovely rereading of some beloved classics from adolescence and an interesting look at why we reread novels and what we gain from so doing at different times of our lives.
- More Than Friends: I couldn’t really empathise with this, but I thought it rather beautiful regardless – about relationships that are very much relationships, they carry weight and they matter and there is love, and lust, somewhere, but they never quite become fully-fledged. “When I think about the singer at all, it’s usually because I had a dream about him. It’s amazing how the details are all still there in my brain, even twenty years later: the rubbed-thin feel of his band t-shirts, the oakmoss notes in his cologne, the way his hair felt on the soft skin on my neck. If we had had sex, I’m sure those memories would be there, too, but we never did. My relationship with the singer exists in a kind of category-less limbo—definitely more than a friendship, but not quite an actual relationship. The singer and I never “made love,” but we did make love, coax it from the air around us, render it in our folded hearts. We made letters and art and songs, we made lists of things we taught each other, we made poetry we exchanged in the middle of the night, walking to the spot exactly between our across-campus dorms, and then walking quickly back in opposite directions. In the winter, he took me as his guest to our college’s winter formal. Our designated driver got drunk, and the singer shelled out for a cheap hotel room across the street from the banquet hall. We draped our fancy clothes across the suitcase rack and slept in our underwear under the stiff hotel blankets, side by side. A thunderstorm raged outside. Lightning flashes filtered through the curtains, throwing shadows on our bare shoulders. He didn’t kiss me.“
- Wife In Reverse: A very short story which I have read five times this week, Time’s Arrow in miniature. Beautiful.
By Pon Arsher
AND NOW, MOVING PICTURES AND SOUNDS!: