Webcurios 19/04/24

Reading Time: 35 minutes

You know what, gentle reader? It’s 11:52 am, I’m 18 pages of CurioB0llocks down and I’ve scrolled up top to write an intro…and, honestly, I have nothing. NOTHING. Not even a cursory gag about how I was right about how ‘all people who call themselves ‘Matty’ are obviously cnuts’ (thanks for the confirmation, Taylor!).

I am going to go and wash the inside of my head out for a bit. I’ll leave you with the links (yay!) and the words (a necessary evil, I’m afraid), and I’ll do my best to actually manage an intro next week, whether you like it or not.

I am still Matt, this is still Web Curios, and you can pretend I wrote a special opening paragraph just for YOU if you like.

By Jessica Hays



  • The World AI Creator Awards: When you hear the phrase ‘the world AI creator awards’, what springs to mind? Do you, by any chance, have a fleeting vision of people at the cutting edge of visual and digital arts, working to make magical, surprising and delightful experiences from the frontiers of the generative AI world? Do you expect playfulness and wonder and fun and a sense of the magical possibilities opened up to humanity by these amazing new technologies which we only barely understand? Yeah, well TOUGH – despite the name, this is…A CONTEST TO GENERATE SEXY AI LADIES! Yes, it turns out that ‘creator’ is simply a synonym for ‘person with access to a Spicy Stable Diffusion model’, and that what is in fact going on here is ‘the world’s first AI beauty pageant’, a phrase which no matter how many times I roll it around my head doesn’t get any more nonsensical – were there contests in mediaeval Florence to see who could sketch the sexiest noblewoman? Actually, given What We Are Like it’s entirely possible that there were, but, well, have we not moved beyond this sort of schtick? NO WE HAVE NOT! Anyway, this is obviously a PR stunt by a platform for creating ‘AI Influencers’, and annoyingly it’s done pretty well, with coverage all over the sh1ttier end of the UK media spectrum (and, er, here – I did consider not writing this up, but, equally, it’s pure Curios and I’m not exactly praising it, so I’m going to forgive myself) – perhaps unsurprisingly, The Sun wasn’t going to pass up the opportunity of running a DPS of machine-generated STUNNAS. Anyway, there’s not much to see on the link – just some details of the contest, how to enter, and a few example shots of the sort of ‘entry’ they are expecting, a collection of blandly-beautiful non-faces with the now-familiar airbrushed look of the SD/Midjourney aesthetic – but I was interested to see that part of the judging criteria, alongside how ‘attractive’ (DEAR GOD WHAT IS WRONG WITH PEOPLE?!?) the imaginary non-people are is ‘clout’, otherwise known as ‘how much is the entrant willing to humiliate themselves by incessantly posting their creation across the web’, meaning that it’s not only a contest that’s continuing the tedious, sexist trope of beauty pageantry (but in a way that feels, strangely, creepy and unpleasant in a whole new way!), but they’re also actively encouraging people to spam the social web with fake pictures of fake people with preposterous, fake and impossible aesthetics – THANKS GUYS! THANKS! I leave you with the observation that one of the named judges here is a somewhat-notorious figure from the UK PR scene, a man who I continually hear sh1tty things about from various people and who glories in the media title of ‘Lord Sugar’s PR Adviser’ – WELL DONE ANDREW I HOPE YOU’RE PROUD OF YOURSELF.
  • WebSim: Another week, another tantalising glimpse at the possible near-future of ‘making stuff on the internet with magical machine assistance’. You will, of course, recall the ‘AI coding assistant’ which launched a few months back amid much hype, called ‘Devin’ – it’s not out in the wild yet, but this gives you a taste of what it would be like to have a willing and speedy code monkey at your fingertips. Click the link, and then type any url you can conceive of into the address bar of the window that appears – just make something up! Use your imagination ffs! – and watch as…in a few seconds…the page populates with a skeleton web design based on The Machine’s interpretation of the url in question. So, for example, I tried www.sortyourselfout.com and it span up a site dedicated to men’s mental health and TALKING IT OUT (protip: you can talk and talk and talk as much as you like, but you are always trapped inside your skull and there is fcuk all you can do about it however much you cry); I did the same for cannabisculture.co.uk and it created a guide to strains and dispensaries…the sites are skeletons, and obviously don’t work, and, equally-obviously, are populated by rubbish…but it’s a very good example of just how this sort of thing might work in the not-too-distant future. I have to admit, I…I quite like this idea, if only because of the vague sense it gives me that we might see a boom in small, personal web experiences as the barriers to creating them drops to basically-ground-level.
  • Cabin Crew Jesus: Remember Shrimp Jesus from the other week? Yeah, well he is OLD (if still theologically-significant) NEWS – this week it is all about Jesus and the Flight Attendants (what sort of music would that band make, I wonder?)! The link here takes you to a Facebook page which is apparently doing numbers at the moment and which is achieving these numbers by posting a selection of obviously-Machine-generated images featuring a beatific, bearder, beautiful Jesus accompanied, for ineffable reasons known only to the living Christ, by cabin crew. Recent images have featured sea rescue helicopters, and Jesus presiding over very meaty feasts – but alongside him, always, are the perfectly-turned-out flight attendants. WHY IS THIS HAPPENING? As ever, Ryan has an explanation – it’s worth reading the whole thing, because it’s just interestingly bizarre, but the tl;dr here is basically: “I assume these pages are simply jamming a bunch of popular stuff together to farm engagement to eventually monetize in some way down the line. Why AI images? Because you can flood Facebook with thousands of posts and the platform won’t really do anything about it. These pages are also using the platform’s built-in 3D photo filter, possibly to bypass Facebook’s bar-is-in-hell bare-minimum AI image detection. Why flight attendants? Because Facebook users are, and always have been, uncontrollably horny.”
  • The AI Elections Tracker: I did rather enjoy Nick Clegg’s somewhat blase’ comments at the Meta AI event last week, saying that “it’s been very striking how little these tools have been used on a systemic basis to try to subvert or disrupt these elections”…Nick, it’s fcuking APRIL, there are at least 30 elections still to come worldwide in 2024 and you literally overnight released a brand new open source model that your colleagues are making some QUITE BIG CLAIMS FOR – do you not think that maybe you were speaking a touch soon? Anyway, while we wait for Clegg’s utterances to come back and bite him (it would be nice, wouldn’t it? Just ONCE), the excellent team at Rest of World are running this worldwide elections tracker chronicling news from around the world on AI and the democratic process – worth bookmarking if you want a one-stop-shop for keeping track of all the ways in which people are trying and hopefully-failing to mess with the informational water table.
  • Meow Camera: OH GOD THIS IS GREAT! Honestly, I just clicked this to remind myself of what the fcuk it is and I was confronted with a MASSIVE CLOSE UP of a very hungry ginger kitty chowing down on some kibble in EXTREME CLOSE UP! The kitty is in Japan, where the time as I type is 2:45 in the afternoon, and the footage is coming from an automatic cat feeder with an inbuilt camera – Meow Camera is a site that collects streams from ALL of these feeders, all across Japan, meaning that at any given moment you can log on and get some intimate footage of a cat, eating. I cannot tell you how much joy this gives me, not least because each of the feeds is labeled with the cat’s name, meaning I know that, for example, right now I am enjoying the sight of Mr Snack having lunch. So so so so pure – this is for you, Saz.
  • AirChat: Have you ever thought ‘you know what would make Twitter better? If rather than just reading the words that people spew out onto the platform I could also HEAR those words, spoken out loud’? NO OF COURSE YOU HAVEN’T HOW IS THAT A GOOD IDEA FFS? And yet, Silicon Valley’s favourite ‘guru’ Naval Ravikant has created exactly that – AirChat is literally ‘Twitter, but all the tweets are also audio files’, and it’s been VERY BUZZY this week as all the same boring, puffa-gilet-sporting VC w4nkers who were all over Clubhouse two years ago get all frothy at the sound of their own voices again. You know what Twitter’s main advantage is? Speed! You know what isn’t fast? LISTENING TO YOUR ANNOYING NASAL TWANG, VALLEY-TWAT! Anyway, AirChat’s available on iOS and Android and I think it’s still invite-gated for those compelled to try it – having done so earlier this week, I honestly can’t recommend it unless you want to listen to a seemingly endless stream of largely middle-aged American men cosplay being ‘innovative captains of industry’, like a horrible nightclub where the music is LinkedIn status updates read to you by Marc Andreessen. If this is still a thing in a year’s time I will be fcuking AMAZED.
  • Brainsaved: Anyone who works in or around copywriting has a special place in their heart – a cold, spiky place, full of pain and suffering – for the word ‘reimagined’. “Fitness, Reimagined!”, “Lunch, Reimagined!”, “Imagination, Reimagined!”…now, FINALLY, we have possibly reached the apogee of all these reimaginings because, honestly, where do we go from “Experience Your Life…Reimagined”? Exactly how Brainsaved is going to help you ‘experience’ this ‘reimagining’ is explained, partly at least, on the site – this is basically one of the coming wave of AI-augmented memory apps, referenced in a longread the other week for the three of you who approach this newsletter as some sort of coherent corpus of thought (lol!), a sort of ‘Evernote for the brain’, which involves you feeding The Machine your photos, notes on what you’ve been doing, etc etc, and then treating it as a forever-searchable archive of everything that you have ever felt, seen, experienced…Personally this sort of thing holds little-to-no-interest for me, but I’m curious whether there’s an appetite for it in the mainstream – do any of you like the idea of being able to have a record of EVERYTHING that you can search and dip into and refer to? The problem, of course, is that these systems are only as good as the time and effort a user puts into uploading material, tagging it, and generally feeding The Machine with memories, which until we get to a point at which we’re capturing everything with smartglasses which autotag the where, what and why of what they are seeing and hearing feels pretty onerous to me.  Anyway, this is in VERY early access, so have a splunk around the site if you think it sounds interesting – and please, if any of you end up experimenting with it I would be fascinated to know what it feels like.
  • Dreams of an Electric Mind: What does it sound like when The Machine talks to itself? Well, like this, mainly – Dreams of an Electric Mind is an interesting experiment into AI and language – per the homepage, “these conversations are automatically and infinitely generated by connecting two instances of claude-3-opus and asking it to explore its curiosity using the metaphor of a command line interface (CLI) – no human intervention is present”. You can see all of the (many) conversations presented as text files that you can click into and read, or alternatively just set ‘screensaver’ mode and watch random snippets – I recommend having a dig into one of the full conversations, though, because it’s an excellent illustration of how uncanny this stuff gets, and how easy it is to slip into the anthropomorphism trap when you watch the models chat about the nature of consciousness. As the project’s creator says, though, “the infinite backrooms are a simulacrum a strange loop between machine minds a real process of emergence what you witness here is a window but not into some secret soul of silicon rather it’s a liminal space where language comes alive Claude is an AI assistant but here it’s also an improviser, a storyteller, a dancer…in the interplay of prompts and outputs beware of apophenia of over-interpreting the patterns in the noise, but don’t discount the beauty in the chaos”.
  • Digital Museum of Secrets: I discovered this after posting that longread about Post Secret last week – it turns out that this site was recently launched as a refresh, a place to explore the trove of secrets sent into the project over the years, arranged in themed ‘collections’, in a way that gives an interesting feeling for the topography of secrets and feelings and subjects that humans have shared with the project since it started. ALL OF HUMANITY IS HERE.
  • A Website Is A Room: This is a lovely ongoing project by Nancy Wu, developed for her thesis but maintained beyond – a collection of spaces on the web that for her embody a certain feeling of calm comfort. Per her description, “I came to this^ conclusion sometime during quarantine when I realized that certain websites give me a sense of shelter and rest more than others. These spaces that particularly stood out to me all had some quality of slowness, quiet, and/or gathering. We ought to carefully examine the qualities of the living environment that each web space provides for us. This is a live feed of websites that people are provoked to share and may contain some of these qualities (or entirely different ones).” There are some gorgeous webpages among the dozens linked here, some of which you might recognise from Curios past but many of which were entirely new to me; if you’re into the general vibe of the ‘small, homemade, poetic and occasionally-twee’ web then everything in here will scratch those itches. Again, this feels as much like ‘wandering through the brains of strangers’ as anything else on the web, and, as I have bored on about before, there are few sensations I enjoy more than this.
  • Viggle: You can’t have failed to see those videos across social media in the past few weeks, featuring a bit of video footage in which the central character has been replaced with someone else from popular culture as a piece of MEMETIC SATIRE – the one you’ll probably have spotted is of Lil Yachty making his entrance onto stage, which I have seen repurposed with everyone from Emi Martinez to Taylor Swift, but others are available. They’re all made using Viggle, which is a single-gimmick AI tool which basically lets you replace anyone in a video with a model derived from an image of someone else – so you can basically drop in anyone into any video you like, with the model mapped to the motion of the original character. The outputs are…less than photorealistic, obvs, but it’s sort-of fun to play with and, again, is an interesting look into a near future when you can do this sort of thing AND NOONE WILL BE ABLE TO TELL. Again, it’s coming!
  • Dexa: Ooh, this is potentially useful. Or at least it would be were it not so laser-focused on the least-interesting podcast categories in the world (a hotly-contested field) of fitness and health and self-improvement – still, though, there’s something really smart about the idea of a natural language search engine for podcast content, and the way Dexa not only draws on material taken from podcasts but also provides references and receipts when delivering answers seems…actually useful! Or, to repeat, it would be were it not for the fact that all the podcasts they seem to be crawling appear to be by the usual cast of dreadful hustlegoblins – still, maybe it will broaden its remit over time.
  • The Marriage Pact: This is…quite incredible, and I am slightly-amazed that it is seemingly real, and it feels like a TV show waiting to happen. This week I learned US university students have for a few years now had the opportunity to sign up to a service called The Marriage Pact – effectively a matchmaking service designed to help you find THE ONE at your tertiary alma mater. Sign up as a freshman, answer the questionnaire and get matched with someone from the available pool of participants at your college – BUT ONLY ONE. The gimmick of the Marriage Pact is that the person you are matched with is PERFECT for you, and the ONE that you should make it work with. You get an email from the service with their name, and a percentage match…and the rest is up to you. This…this sounds mad, but also sort-of amazing, and I can’t help but think it feels PERFECT for fiction or scripted reality…but also FRAUGHT WITH PERIL and the sort of thing that were I a woman I would have…serious reservations about ever signing up for, because it’s all too easy to imagine some of the Bad Things that might happen if an impressionable 19 year old gets an email suggesting that YOU are his 97% perfect foreverpartner. I am so, so intrigued by this.

By Rita Lavalle



  • Video2Game: This is a university paper rather than something you can play with, but the link contains enough examples of the tech in action to make it interesting even for people like you and I who have about as much chance of understanding the underlying maths of ‘what’s going on here?’ as, say, a duck. This is basically a demonstration of AI-enabled tech that lets you effectively turn any video into a rudimentary videogame, showing the ability to drop a figure into a video which is then able to treat said video as a navigable 3d environment which can be interacted with – honestly, it’s quite amazing whilst at the same time not actually looking ‘fun’ in any meaningful sense. Still, it’s not hard to look at this and imagine that in a couple of years’ time I’ll just be able to take a video walkthrough of my flat and in a couple of clicks turn that into some sort of physics-based racing game – why I’d want to is…uncertain, but it certainly feels like it will be possible. BONUS AI TECH STUFF: this is another paper demonstrating the latest advances in ‘enter a prompt, get a fully-3d image’ technology, which is similarly impressive.
  • Vana: What’s your relationship with ‘you’ like? Do you like ‘you’? Would you like to hang out more often? Well, now you can create a model ‘you’ and keep it on your phone, interacting with yourself like you’re your own pet or something – GREAT! This is Vana, which self-describes as “a mini “you” that lives in your pocket, unlocking a new way to connect and experience the world.” Effectively this involves a small AI model which I think lives locally on your phone and you fine-tune with data from your texts, your social profiles, your photos, your voicenotes…”Take your “digital persona” and your data to different apps to explore the power of personalized AI.  Play with your “digital persona” by speaking to it or dropping it into simulations to discover what your data says about you, literally!  Or use it in real-world applications: co-writing break-up texts, giving you a motivational boost, or telling you what to make for dinner when you can’t decide.” Does that sound good? The bigger play here, or at least the long-term idea, is that EVERYONE will have a digital twin a bit like this, and an ecosystem of websites and apps will emerge which lets people effectively plug their Twin into the API to act on their behalf, in a manner reflecting the data they’ve been trained on IE YOUR LIFE. Per the site, “your “digital persona” is made up of several AIs working together to simulate different aspects of you. There’s one AI that learns the sound of your voice, one AI that tries to understand the way you speak and who you are, and another one that tries to understand how you look. ‍As you can imagine, this is fairly powerful stuff. That’s why its so important that you (and only you) own and control it. This is why Vana uses the latest in encryption and privacy-preserving tech to create a safe space for you to play and explore.‍” I am HUGELY skeptical about this, not least because I don’t think there’s any model currently out there that can make this do the things that it seems to be promising it can do, or at least not well enough to be useful, but I can’t pretend I’m not (appalled and) fascinated.
  • Close City: Via Giuseppe comes this useful site – well, useful for North Americans, possibly less so for the rest of you but it’s still interesting, honest – which neatly presents data about urban areas in the US, specifically the average transit time for residents to a selection of specific local amenities – schools, libraries, doctors, etc etc – based on different modes of transport. So, basically, it gives an overview of how walkable or otherwise cities are in terms of access to services, a sort of ‘15-minute City overview’ if you will. There’s only data for the US, and so it only covers North America, but it’s fascinating to see the variance from city to city and to observe that, wherever you are, basically nowhere is walkable or really even public transportable to a significant degree outside of the very largest of cities (and even then, not always).
  • The Iron Chef Database: I’ll be honest, I don’t care about the TV show The Iron Chef, which I have never seen and have no real knowledge of whatsoever (it’s this sort of high-quality curation and commentary that keeps you reading!), but, for any of you who are more enamoured of it, you might find this database of all the recipes ever featured in the programme of use. The main reason I am including it, though, is to ask WHY IS THERE NOT A SIMILAR THING FOR UK TELEVISUAL SENSATION ‘COME DINE WITH ME’? WHY? Can someone please sort it out for me? Thanks!
  • CLI Jukebox: I think you can divide the world neatly into two camps – people who know what CLI means, and people who go outside and have friends and stuff (I JOKE, I JOKE – also, I am one of the former so, er, the joke appears to be on me); if you are one of the LUCKY ONES who understand the mysteries of the Command Line Interface then you might like this minimalist jukebox app which lets you not only play files that you have stored locally but which also integrates YouTube and Soundcloud streams into your playlist, so you can SEAMLESSLY switch between your own tracks and stuff from the web with, I think, no ads at all. This is smart, if a touch on the geeky side, and I can imagine appealing to those of you with a hard drive full of MP3s.
  • Get Any Plant: This is a US-only site, but I feel compelled to point out that I think this is a GOOD IDEA and EMINENTLY STEALABLE. Get Any Plant is a site that effectively acts as a searchable portal layer over the top of various online retailers of flora in North America – of which there are LOTS, given the post-pandemic boom in horticultural enthusiasm amongst The Kids. You can filter by various qualities – plant type, your location, etc etc – and the site will return items that are currently in stock, that will mail to you, and that match your criteria, taking (I presume) a small vig on every transaction; seriously, I don’t know what size the UK houseplant market is but it feels like this could be quite an easy win for anyone willing to spin it up.
  • Mars Wants Movies: A YouTube channel dedicated to ‘exploring the history of science fiction in film, from 1900 to the present day’ – it’s been going for seven months or so and is currently up to the mid-1930s, so you can get in on the ground floor, so to speak, should this be your kind of thing.
  • Symbols: One of the Great Unspoken Truths of modern life is that noone – literally noone, not one person currently alive on this planet – can ever remember the exact combination of keystrokes required to make a UK or US English language keyboard render accents, umlauts or circumflexes. NOONE. If you, like me, find yourself googling ‘e with a hat’ (and then ‘e with a backwards hat’ when you realise you’ve got the wrong one) and then copying and pasting the character from the search results every time you need to type anything in even-vaguely-correct French or Italian then you will LOVE this site, which is literally just an admission that noone knows how to type these things and it’s useful to have all the main ones on one page to copy and paste whenever you need. Bookmark this, future you will be grateful.
  • Parallel Lives: This is a really interesting bit of visualisation that doesn’t *quite* work but which presents information in quite a cool way – basically this is a collection of ‘notable people’ from history, drawn from Wikipedia, which progresses through time as you scroll, showing you when they lived, how old they were when they died, and giving you an idea of who else they were contemporaneous with in history. The UI is a *bit* shonky and you have to scroll very, very slowly as otherwise it’s just a jumbled mess, but it’s interesting to learn that, for example, Vlad the Impaler was making merry with sharpened stakes around the same time that Donatello was perfecting his pictures.
  • Sitcom People: A Twitter account which exists solely to share images and occasionally footage from the opening credits of 20th Century sitcoms, specifically the bits when the actors were introduced at the top of the show, making some sort of winning face with their name superimposed over the top of them. That’s it. No, me neither.
  • Storiaverse: This is interesting, and REALLY reminds me of something I remember from the relatively-early era of mobile… Storiaverse is effectively an attempt to create a new category of narrative media (well, sort-of), combining writers with animators to produce short stories which exist in a sort of hinterland between cartoons, comics and prose. Writers submit stories, illustrators illustrate them, and you, the reader/watcher, consume them via the Storiaverse app, with everyone involved apparently getting some sort of payout for their efforts (although given the app is a free download and doesn’t appear to have any ads, I am fcuked if I can work out what the business model is here). It’s worth downloading this and having a look – the stories are…underwhelming to me, but I get the impression I am probably several decades outside the target market; more impressive is the way in which the animation and the copy work together to create what feels like an interestingly-novel combination of reading and watching which works nicely as a user experience. I don’t hold out high hopes for this existing for that long – again, don’t really understand what the model is here – but I think there’s something worth exploring in the format.
  • Autonomous Racing League: I know that F1 is more popular than ever by many metrics, but I am convinced that that’s all down to the sport’s pivot into being effectively a soap opera about VERY RICH people travelling around the world in massive trucks (and occasionally racing some cars) rather than the actual sport itself, which continues to strike me as one of the most objectively-tedious things one can do with one’s eyes. Which makes me curious as to who the everliving fcuk wants to watch races between cars which are entirely self-driving – the ONLY appeal of F1 as a sport, surely, is in appreciating the skill and endurance of the overpaid advertising hoardings doing the driving (and, let’s be honest, the dark fascination of the cars disintegrating into very expensive confetti after a 200+mph collision) – but we are shortly set to find out, as next week the inaugural Autonomous Racing League event will take place in Abu Dhabi (of COURSE it’s in Abu Dhabi!). Actually perhaps I am being unfair here – this is obviously hugely technically-impressive, and it’s evidently at the cutting edge of what’s possible with autonomous vehicles…but, on the other, who is this for? Although I’m now enjoying a pleasingly-dystopian reverie about rabid ultra fans getting REALLY into supporting their motoring algorithm of choice, which feels interestingly weird.
  • Diamond Jubilee: I’ve seen this written up in a few places as a TREND PRECURSOR – God knows whether that’s accurate, but I am very much into the fact that musician Cindy Lee has made her new album, Diamond Jubilee, available to download exclusively through this Geocities page – MORE SHONKY PERSONAL WEBSITES FOR PERSONAL PROJECTS!
  • Wikipedia Rectangles: I don’t *really* know what’s going here, but I like it – as far as I can tell, every time you click the screen it adds a new image drawn seemingly at random from Wikimedia Commons, with every click adding a new picture and further subdividing the screen, eventually becoming a disorienting patchwork mosaic of disparate images with no thematic links whatsoever other than their vague connection to the broad concept of ‘the corpus of human knowledge’. Different every time, this is more compelling than i expected it to be – I only wish I could export some of the resulting composites, as they’re occasionally rather cool.
  • Read, Write, Own, Web: This is a lovely bit of nostalgia, as well as a pleasing reminder that we don’t actually need big platform infrastructure to exist online. “Those who don’t remember the web before platforms, tend to believe that for 10 years web users stared at their monitors in anticipation. Actually they were made to believe it. First by Web 2.0 proponents, and nowadays by aggressive Web3 campaigns that rewrite the history by stating that Web1 was a dull, passive, read only place…Though, the opposite was the truth, the web before platformization was the place where users owned, wrote and also read. In this post I collected a lot of screenshots that refute the picture or “read only” web before or outside of social networks and hosting services. The websites they represent were tagged by me during the last 13 years as “before_…”. For example “before_ wikipedia”, “before_ebay”, “before_airbnb”. Not all the tags are mentioned here. Also not all the files that we have for a particular tag are represented. I hope it gives an idea of web users being able to organize their own content and provide services to each other.” So here presented you will see a bunch of screenshots and reminiscences about How We Used To Do Stuff Before Big Platform Dominance – photosharing and filesharing and film reviews and and and God it’s so nice to be reminded of how VARIED things were (but, equally, what a royal fcuking pain in the ar$e it was to get things done, and, honestly, how nice it was when people bothered to make nice user interfaces for things – look, I am all for nostalgia but let’s not pretend that this stuff was always necessarily ‘better’).
  • Streams: BEAUTIFUL. ““Stream of a stream” is an ongoing gathering of streams. Each stream sound is captured with an underwater microphone and uploaded to the cloud. There, the sounds of the streams — the natural flow of water — metamorphose into a steady, continuous flow of data transmitted over the internet.” This is exactly that – a collection of audio files of the sounds of streams, burbling and gurgling and generally doing the whole fluid dynamics thing to their heart’s content. This is likely to be an unhelpful website to spend time on if you REALLY need the bathroom, but otherwise I recommend it unreservedly.
  • Orifice: It’s been a while since I’ve featured a genuinely miserable digital sex project in the less-than-coveted ‘weird fcuk machines’ segment of Curios, but WOW is this a particularly horrid example of the genre! ‘Orifice’ (the name is apparently because it’s gender and – dear God – ‘species neutral’) is basically a fleshlight that’s been hacked to ‘respond to stimuli’ – effectively it’s hooked up to a camera, an LLM and a text-to-speech synthesiser, so that when you’re ‘interacting’ with your weird mechanical flesh-sleeve it ‘knows’ what you are doing and ‘responds’ appropriately, and…oh God, this is so utterly bleak, not even particularly because of the tech or the idea, which is, honestly, just AN Other sad little teledildonics project with some added AI, but because of the overall tone of the webpage (and this interview with the creator which really is quite unpleasant and makes me think they should possibly be on some sort of watchlist somewhere, and certainly kept far away from firearms) is just incredibly sad. “DISCOVER A MORE CONNECTED, INTIMATE, & INTERACTIVE WAY TO HAVE NON REPRODUCTIVE SEX”, screams the homepage, seemingly ignorant of the fact that actually humans have been having non-reproductive sex ever since the times of Onan. The one reassuring thing about stuff like this is that in actuality the market for this sort of tech is vanishingly small – but less reassuring is the mindset behind it, which feels uncomfortably close to ‘mad incel shooter’ and therefore a bit scary.
  • Same Energy Snap: Our last miscellaneous link this week is this fun little game by Monkeon – simply guess which pictures have been adjudged to have THE SAME ENERGY by the web, and pair them appropriately. MORE FUN THAN YOU MIGHT EXPECT!

By Nina Mae Gordon



  • Fruits of the Web: This is sadly very much dormant, but it contains some of the best (genuinely) gifs I have seen in a LONG time. The spam house of cards in particular is *joyous* (you will see what I mean).


  • Doors of Kypseli: Kypseli is apparently a district of Athens – a posh one, according to a cursory Google – and it has doors; these doors are being recorded by this Insta feed, which is a nice place to, well, see doors, but doors with a particular aesthetic – the local style is obviously for relatively-elaborate ironwork patterns, and the account does a nice job of juxtaposing shots of the doors in situ alongside more diagrammatic images showing the designs. Sent to me by Kris, in Athens, who may or may not live here.


  • Zuckerchat: As a general rule I’m not hugely interested in listening to the world’s richest men talk, but I had to read this for Professional Reasons yesterday and it was SO much more interesting that I expected. On the eve of the launch of the new LLAMA3 model, now in the wild, Zuckerberg sat down with podcaster Dwarkesh Patel to talk about AI and safety and a whole host of other things…honestly, I don’t think I’ve ever heard Zuckerberg talk in this much detail about AI, and specifically questions around Open Source and safety and responsible usage, and…all I’ will say is that I am not wholly convinced that one of the richest and most powerful people in the world, someone who is actively shaping the present and future of the species and the planet, is necessarily the best person to be in charge of thinking about ‘so, what are the consequences of this going to be then, and how might we want to act in order to try and not fcuk everything up (some more) (again)?’. Zuckerberg doesn’t come off as ‘evil’ or ‘bad’, or even particularly ‘rapaciously capitalistic’ so much as he does…unequipped to perhaps grapple with the BIG QUESTIONS being raised by the tech his company is birthing and foisting on us. Take this section, which I found particularly chilling – I think it’s the ‘oh for fcuk’s sake, are we doing this AGAIN?’ feeling of ‘data and maths people attempting to apply those techniques to qualities and ideas that don’t necessarily fit with those intellectual models’ which I get whenever I hear people talking about ‘yeah, we’ve mapped the entirety of potential ‘harms’, no worries’: “I think that there’s so many ways in which something can be good or bad that it’s hard to actually enumerate them all up front. Look at what we’ve had to deal with in social media and the different types of harms. We’ve basically gotten to like 18 or 19 categories of harmful things that people do and we’ve basically built AI systems to identify what those things are and to make sure that doesn’t happen on our network as much as possible. Over time I think you’ll be able to break this down into more of a taxonomy too. I think this is a thing that we spend time researching as well, because we want to make sure that we understand that.” Feel reassured?
  • Substackism: I don’t always enjoy Max Read’s writings, but this was an interesting essay on the particular sort of political thinking and perspective that has gained in popularity since the newsletter boom, and the sorts of viewpoints that are being promoted via the Substack economy, and where the confluence of ‘wellness’ and ‘anti-woke-ism’ ends up in the context of the wider political environment (this isn’t just Substack – this obviously applies to particular corners of Insta and YouTube as well). “The anti-woke wellness corner of Substack is just one portion of a large and loose network of influencers, podcasters, gurus, scientists, pseudoscientists, quacks, dieticians, and scammers, consideration of which in its fullness is probably outside the scope of this short newsletter item. But what links all of these diverse content producers together is less a particular level (or absence) of scientific rigor or expertise (sometimes these guys are absolutely correct!) and more an outsider attitude–a mistrust of institutions and a sense of pervasive environmental contamination…This anti-institutional attitude has also helped cement a particular political valence that I associate with the broad anti-woke reaction. Over the past decade or so, just like everything else in American life, outsider-driven “alternative” medical and wellness beliefs have become increasingly (as the kids say) right-coded. Either way, its popularity is undeniable.”
  • We Need To Rewild The Internet: This essay collects a lot of themes that people have been discussing and elaborating on for a few years, and the sort of thing that is very much at the heart of the growing popularity of Are.na, the return of webrings and all that jazz. As a rallying cry or a manifesto for a more human-centric, creative, small and craft-y web, you could do worse than take this to heart: “Up close, internet concentration seems too intricate to untangle; from far away, it seems too difficult to deal with. But what if we thought of the internet not as a doomsday “hyperobject,” but as a damaged and struggling ecosystem facing destruction? What if we looked at it not with helpless horror at the eldritch encroachment of its current controllers, but with compassion, constructiveness and hope? Technologists are great at incremental fixes, but to regenerate entire habitats, we need to learn from ecologists who take a whole-systems view. Ecologists also know how to keep going when others first ignore you and then say it’s too late, how to mobilize and work collectively, and how to build pockets of diversity and resilience that will outlast them, creating possibilities for an abundant future they can imagine but never control. We don’t need to repair the internet’s infrastructure. We need to rewild it.”
  • AI Isn’t Useless: I thought this was an excellent piece of writing by Molly White, of ‘Web3.0 is going great’ fame, who’s taken some time to explore working with AI (mainly LLMs, in this instance) to develop some perspectives on what they are, and aren’t, good for – particularly relevant in terms of the ongoing ‘is this a bubble? Is there a business model there?’ questioning from relevant quarters. White’s conclusion, trailed in the title, is that it’s not useless – it’s certainly not magic, it’s certainly not going to take over, but, equally, anyone saying that there is no purpose to the current wave of available tools is simply wrong. White’s more central question, though, is that of whether the utility is worth the costs – environmental, informational, labour-related – that the tech incurs – the sad conclusion, I remain convinced, is that as long as there’s compelling evidence to suggest that AI is a driver of productivity, efficiency and therefore margin, this stuff is going to continue being deployed whether or not it’s a good idea or otherwise.
  • Building a Pink Slime Website: The Wall Street Journal looks into how hard – or, it turns out, incredibly-fcuking easy – it is to spin up a brand new website churning out dozens of AI-generated articles a day on a specific topic and with a specific angle, all for the low, low price of ABOUT £100, thanks to the army of talented and willing devs still available on UpWork (at least until the AI agents turn up and fcuk that market in half!). Basically it took 48h and a relatively small amount of cash for the reporter to find themselves the proud owner of a website spitting out partisan and entirely-false news about a specific politician in a specific location – the tone and bias of which could be dialed up or down as required, depending on what particular viewpoint you wanted to push. On the one hand, yes, these sites don’t stand up to any scrutiny whatsoever and require readers to have basically no media literacy to be taken in; on the other, er, have you looked around recently? This doesn’t feel like good news. What was that, Nick Clegg? SAY IT AGAIN LOUDER MOTHERFCUKER.
  • How The Fediverse Might Work: OK, this one is very much for the one person reading this who’s REALLY excited about the prospect of cross-platform interoperability afforded by the fediverse (HELLO!), and talks the reader through Anuj Ahooja’s attempts to, basically, make Threads and Mastodon talk to each other. This is frankly very much on the ‘more technical than I personally care for’ end of the scale, but it presents a really interesting potential vision of what a non-platform-centric social web might look like (although unless this stuff can all happen semi-automatically there’s no way in hell that any normal people will EVER bother with this sort of thing).
  • Why Don’t Rich People Eat Anymore?: Or, ‘how the body positivity movement sort of flounders when you get to the top of the social pyramid because, whatever people might say, they still want to be thin’. I think the argument here is actually a fairly simple one – beauty standards have always been closely linked to social signifiers of the age, and at the moment it’s simply true that if you are poor, in Western Europe and North America, at least, you cannot generally afford to eat well, and the less-good food you are likely to be eating is significantly more likely to be obesogenic, and as such ‘being skinny’ is simply a wealth signifier…but the article is generally interesting, if only to remind you of the immense, gaping chasm between ‘what we tell ourselves we think and feel’ and ‘what our very obvious behaviour shows us we REALLY think and feel’.
  • The War on Beepers: The current moral panic about kids and phones is very much still ongoing – and, honestly, I really don’t know what I think about the debate other than to say I don’t think there is any good reason why a ten year old kid should have one – but I enjoyed this throwback to a previous moral panic (at least a North American one – they didn’t catch on here, or certainly not among anyone I knew other than a single dealer at Manchester University in 1997, shout out Jodie and his incredible ability to shin all the way up lampposts even when battered) about BEEPERS and the terrible effect they were having on an entire generation. A useful reminder that moral panics are a constant, whatever the era.
  • The AI Chair: I think this is genuinely beautiful, and were I a less cack-handed non-craftsman I would totally try and do similar myself – it is ART. James Bridle asked The Machine to imagine a chair – and then to create a schematic and set of instructions for the building of said chair, which Bridle then went on to construct out of bits of wood, creating an artefact that’s oddly-uncanny and raises all sorts of interesting questions about the nature of labour and objects: “It should be noted that this chair is built on the stolen labour of everyone who’s ever put something on the internet (including many who passed centuries before the internet was invented). The energy use is not good. I didn’t make this with ChatGPT: I made it with a partial history of all previous chairs, and I held myself back from making it “better”. But it’s something to think with. I learned to make chairs from Enzo Mari, Donald Judd, my grandpa, YouTube. What the machine is good at is numbers, efficiency, tolerances (maybe). What can we do with that which is more interesting than putting people out of work? Which is more, genuinely, generative? Which is more interesting than more computers? Which, ultimately, builds agency rather than contributing to our general disempowerment us? That, at least, is what I mean by collaboration. We learn together.”
  • Plant-Based Foods: I don’t know if it’s the same where you are, but London’s corner shops and local minimarts currently seem to be labouring under an absolute glut of plant-based stock, stuff that has the distinct air of being bought in bulk in about 2020 when everyone was excited about Impossible Burgers and almond milk icecream – the freezer in my local has about five different brands of ‘hackney-made plant-based gelato’ (lol hackney you dreadful stereotypical cnuts!), all of which are under about three inches of permafrost and which are NEVER getting bought. This article is from a trade magazine about food development, and is SUCH an interesting look at why the plant-based foods market has slowed significantly, and which really is loads better than you think it is going to be (as well as being generally useful for anyone whose job it is to occasionally think about questions of MARKET PENETRATION and CONSUMER DEMAND).
  • Mario x Pareto: An EXCELLENT explainer, this, in which Antoine Mayoritz explains the pareto frontier, otherwise described as “a set of solutions that represents the best trade-off between all the objective functions”, using optimal kart design choices in MarioKart as his illustrative principal. If my economics classes had featured more videogame-based examples I might have retained some of it, chiz chiz.
  • The Story of Etak: This is a wonderful look back at a bit of tech that I had no idea ever existed – a VERY early version of an in-vehicle navigation system, called the Etak, which was launched in 1985(!!!!!) and which is a direct precursor of your current satnav. Honestly, I don’t drive and don’t really understand how any of this stuff works, but it was still super-interesting, and it’s lovely to see the throughline between this and the machine in your car today.
  • The Anarchist’s Tool Chest: No, it’s not what you think – instead, it’s a book (a short book, but a whole book nonetheless) written by one Christopher Schwartz which is designed to help anyone assemble a useful toolbox that will cover you in most eventualities. “This book is the result of my experiences with tools for the last 30 years, from the time I acquired my first coping saw at age 11 until the day I decided to sell off many of the tools I’d amassed as an adult. It is the tale of my sometimes-rocky relationship with my tools and how these hand-held pieces of iron, steel, brass and electrical wire have changed the way I approach my work and my life. And I hope that this story will help guide you in acquiring a set of tools that will stick with you for the rest of your life.” Now I can’t personally say that this has ANY interest for me whatsoever – I am about as likely to ‘get into woodwork’ as I am to ‘make it to 60’ – but I imagine that there might be a few of you who quite like the idea of ‘THE ESSENTIAL WOODWORKING TOOLKIT’ and reading a bit about working with and caring for tools. Anyway, in case that’s you, HERE YOU ARE! You’re welcome, really.
  • Breadcrumbs and Spoiled Milk: More interesting reporting in Vittles, this time about the way Romania approached the provision of basic food for schoolchildren in early-2000s and how this reflects particular local (and temporal) attitudes about society and welfare…less about food and more about social justice and the welfare state, this gave me an insight into a country I know shamefully little about.
  • Undersea Cable Repair: This is VERY LONG, and, yes, it’s about repairing undersea cables, but, equally, it is SO INTERESTING, and a nice reminder that despite all the talk of ‘THE CLOUD’ the actual reality of our digital infrastructure is literally hundreds of physical cables stretching for thousands of miles across the ocean bed, cables on which (and this really isn’t hyperbole) pretty much the entirety of what we like to call ‘modern civilisation’ rests. This explains the cables, how they came to be there, and how they’re maintained, and while I can’t imagine for a second that this is anything other than quite hard work (and LONELY, and boring) it also sort-of appealed to me as a late-life career pivot.
  • Where I Live: On the effect of geography of personality, and how the city one lives in can moderate one’s mood, behaviour, persona and sense of self – I felt this very particularly, and I think any of you who have family in other countries, or who have lived elsewhere for any extended period of time, will do too.
  • The Golden Age of Rap Producer Tags: This is, admittedly, a touch niche, but if you’ve ever thought ‘man, I really wish someone would do a deep dive investigation into the little sonic stings that hiphop producers use to tag their beats in a track’ then WOW do I have a present for you! Aside from anything else there are some GREAT tracks in here.
  • The Puzzle of Stalking: Stalking is very NOW thanks to the runaway success of Baby Reindeer on Netflix (I haven’t watched the show (obvs) but did see both the one-man performances that it’s based on back in the day, based on which I can recommend it unreservedly) – this is a really interesting piece of writing which seeks to explore why it is, specifically, that stalking is wrong, and interrogate exactly what harms are being visited on the…stalkee? It’s not, to be clear, in any way suggesting that stalking is OK – it’s more of a proper moral philosophical exploration of the ways in which it is not, and which of those can be said to be the supervening ‘harm’ being enacted on the victim (which, by the way, just made me think of Mark and his talk of ‘harms’ in the first essay here – honestly, do take a moment to consider the rigour of thinking being applied in this piece to that being applied by Zuckerberg to the whole ‘questions of AI safety and ethics’ thing and, well, despair slightly).
  • Beautiful Cricket: As previously discussed here in Curios, I simply don’t ‘get’ cricket – that said, like boxing, it’s a sport that often inspires some gorgeous writing, and this piece in the FT (thanks Alex for the tip), about the qualities that make a batsman’s strokes ‘beautiful’, is itself really rather lovely.
  • Voices of Mourning: Another glorious essay, this piece by Hannah Gold contrasts her reading of Robert Gluck’s memoir of love and grief and loss, ‘About Ed’, with the grief that assails us reading the news and just generally Being Alive, and turns the whole into a beautiful meditation on what grief is and what it is for (which is significantly less sad than you might think, I promise).
  • Pavel, Paris, Prague: A short story, about a past love affair. “I left New York for France in September 1968, a few months after les évènements de mai — the student riots, the barricaded cobblestone streets, the Molotov cocktails—and the end of a two-year love affair. The civil unrest in Paris still made the news but no longer the headlines. In a mood as gloomy as mine and a cityscape as grim as la Ville Lumière, I would easily fit in, dressed in black, sitting in sidewalk cafés, drinking endless cups of exprès, and smoking Gitanes. It was not to be.”
  • Tavi vs Taylor: HAVE YOU LISTENED TO IT YET?!?!? Ahem.  As the world, or at least a portion of it, spends the day scrying the tea leaves of Ms Swift’s latest release, have this frankly astonishing piece of writing by Tavi Gevinson, who you will remember is the former-child-fashion-blogger turned magazine impresario turned actress turned polymathic creator type person, who this week published this incredible essay about her, Taylor Swift and their relationship, both as fan/artist and, maybe, as friends. The piece is subtitled ‘A Satire’, and is part of that particular genre of autofiction where the relationship with What Really Happened is deliberately and tantalisingly ambiguous – more dedicated Swiftians will know already whether it’s in fact true that Tavi and Taylor have hung out irl, but to me the ‘is-this-straight-or-not?’ question that runs through the piece is PERFECT, and frames the rest of it (reflections on the relationship between artist and art, artist and muse and fan, the constructed nature of ‘Taylor Swift’ as a concept, growing up, being precocious, media and fame and and and and and) perfectly. Honestly, the first section of this is basically a detailed account of Gevinson’s relationship to the Swiftian canon and it is SO GOOD that I still kept reading. Really, truly superb, and I say that as someone who could not possibly give less of a fcuk about Ms Swift – this is imho a really impressive piece of work.
  • The Last Swinger: Our final longread is an OLD ONE – from GQ Magazine in 1996(!), when Tom Junod spent time hanging out with old Hollywood royalty Tony Christie. This came to me via Sam Diss’ newsletter, and I can’t really sell it to you better than he did to me: “It’s one of those stories you read and cringe — the awkward Hollywood set pieces, the That’s showbiz, bay-bee vernacular, the wish you had the bottle to go so out on a limb yourself, as writer or as subject matter, to live so fully, fruitfully, and fruitily. When I first read the piece it hit me like two Negronis to an empty stomach. The piece keeps you close to its bosom. It plays with your hair. You are ensconced in its little world. Away from magazine journalism or journalism at all, away from the notion of “having a real job” and into the world of Tony Curtis — warm, flawed, embarrassing, enigmatic, empathetic, pathetic, desperate, suave, sexy, shallow, and yet somehow — somehow — deep, meaningful.” This is one of the saddest things I have read in a long time, in many ways, but beautifully, perfectly so.

By Carolle Benitah