It’s Friday, it’s lunchtime, it’s practically the holidays and I just bought Arab Strap tickets – ALL IS RIGHT WITH THE WORLD!
It’s not, of course – everything’s a benighted mess and liable to get worse before it gets better or, more accurately, we all die! – but let’s pretend otherwise as we, the Web Curios family (incestuous, dysfunctional, sickly, genetically compromised), come together for a rare moment of collective joy and hope.
*PAUSES*
Wasn’t that nice? Eh? Oh.
Still, like I care – this is the last Web Curios before the traditional Paschal break, and I am very much looking forward to not paying any attention to A-fcuking-I and associated topics for two weeks (by the time I come back I expect you all to have been replaced with automata). I hope that you have a pleasant fortnight, and that those of you who celebrate the cruel and bloody execution of a poor carpenter who was JUST TRYING TO BE NICE TO EVERYONE FFS have a wonderful, chocolatey time.
I am still Matt, this is still Web Curios, and here’s what crucifixion looks like if you’re curious.
By Steve Kim
THE SECTION WHICH WOULD LIKE TO TELL ANY OF YOU WHO WERE CONTEMPLATING GOING TO SEE ‘BERLUSCONI, THE MUSICAL’ IN LONDON AT THE MOMENT THAT YOU REALLY SHOULDN’T BOTHER, PT.1:
- Create Real Magic: I don’t like to start a newsletter with a let-down, but I feel it important to warn you upfront that, unless your definition of ‘real magic’ encompasses ‘poor-quality branded imagery from a globally-notorious polluter and purveyor of carbonated sugarwaters’, you are probably not going to feel the SURPRISE AND DELIGHT that the title here might promise you. Have you been waiting anxiously for the first sighting of an AI-powered consumer activation in the wild (and have a more miserable concatenation of words ever been written by a human in the course of our species’ long and inglorious history?)? YES YOU HAVE! So, then, welcome to this GPT/DALL-E powered marketing campaign by Coca Cola, which brings together a couple of different AI toys (natural language inputs, text-based image manipulation) to enable users from around the world to remix the Coke visual brand back catalogue into new and exciting and MODERN visuals thanks to the POWER OF ARTIFICIAL INTELLIGENCE! The gimmick here is that you get access to a whole host of old Coke graphic design materials which you can incorporate into your own, ‘collaborative with the machine’ canvas – so, for example, pick a cheery Coke santa and then get the AI to imagine the dystopian scene around him (“Santa laughing merrily and lifting a coke to his lips while the bombs rain around an irradiated London”, that sort of thing) – all the images will be assessed in some way, with the ‘best’ ones being used in Coke’s global marketing, billboards, etc. Which, to be clear, is EXACTLY the same mechanic brands have been using to get cheap/free work out of creatives since the advent of the internet – except in this case they can’t even claim to be offering ‘exposure’ to graphic designers! Oh, and good luck actually getting the site to work properly – you need to sign up, which process is unpleasant enough in itself, and once you do there appears to be a ⅓ chance that the site just won’t work, and then once you’re in the interface is poorly-explained, and a bit clunky, and just not that much fun to play with. What does this show? It shows fcuk all really, but if I were to try and draw conclusions it’s that adding a spurious ‘AI’ layer to what is a tired, hackneyed promotional mechanic doesn’t make said mechanic any less tired and hackneyed.
- AdCreative AI: A few months ago I featured a site which purported to be a one-stop-shop for AI-generated creative work, but which in fact turned out to be a bait-and-switch to promote Wunderman or some other agency – this, though, is seemingly the real thing. “Generate conversion focused ad creatives and social media post creatives in a matter of seconds using Artificial Intelligence. Get better results while saving time”, burbles the site copy! “Simply tell our Artificial Intelligence your Target Audience, the platform you are creating the ads on, it will select the best tone and length for the platform while focusing on your target audience’s pain points.” This is priced VERY keenly – $77 a month for 50 ads a month, from what I can tell, though other tiers are available – and, look, while this isn’t going to produce anything other than the most baldly-functional of Insta/FB/etc ads, and whilst it’s not doing anything that you couldn’t do yourself by cobbling together a few AI services yourself for about half the monthly fee, and whilst I know none of YOU need this and none of YOU will feel the hot, heavy breath of The Future breathing down your professional necks, spare a moment to think of all the people in the Philippines and similar parts of the world who make their living with a Fiverr account and a Canva license and who are going to see things like this whittle their likely income from low-rent piecemeal digital work dwindle to not very much at all. Oh, and the data-cleaners too, per this article. Oh, and the journalists. I don’t mean to keep banging on about this, and I promise I will shut up about it soon, but I really don’t think anyone’s taking the coming jobpocalypse quite seriously enough (or maybe I’m just projecting my increasing unemployability, who knows?).
- The Future of Customer Services: This is a link to a two-tweet thread, but I promise it’s worth the click if only to get a small window into the glorious future that is set to be AI-enabled customer service interfaces. Steve Guntrip was accidentally granted access to a trial being run by photo equipment shop DigistoreEU of a new GPT-powered customer service chatbot – as you will see from his screencaps of their interaction, it does not appear to be a significant upgrade on a human (although the amusing digression into obscene poetry and milkshake recipes was, if I’m honest, a pleasant surprise and the sort of thing I could maybe get behind next time I’m in an interminable menu deathloop on the phone to HMRC. Now, obviously this setup is a test, and wasn’t meant to be live, and is a VERY early days implementation of tech that is only going to get better…and yet, it doesn’t take a massive effort of will to conceive of a not-too-distant future in which all our interactions with the large corporations and cash-strapped public institutions that mandate our lives are managed by systems like this, and exactly how fun it will be when they go wrong but there’s no way of escaping from the horrible recursive AI-led conversation loop and there’s no killswitch to get straight to a human operator and so you’re just forced to listen to endless iterations of “I can’t help you with that, Dave” until whatever is ailing you finally becomes terminal.
- A Small AI-Enabled Desk Pet: By way of antidote to the…less-than-sunny tone of the last few links, here’s something that looks genuinely fun and which feels like another window into a future that’s a bit more whimsical and ludic and playful. This Twitter user has cobbled together a bunch of different systems to effectively create a little holographic cartoon rabbit desktop companion (it’ll make sense when you click, promise) which she can interact with through voice commands and which is obviously running on some local LLM and it is, honestly, SO CUTE. I really think that someone somewhere is going to make a killing selling things like this – but, also, if you’ve read any contemporary scifi in the past few decades (or watched enough B***k M****r), that there will be some…interesting examples of them going rogue. Whilst I am well aware that there are a million-and-one brand safety and protection reasons that mean this could never happen, part of me rather likes the idea of a brand offering an open-source, downloadable AI-enabled version of their mascot that any hobbyist can use to create something like this.
- FreedomGPT: Except, of course, what happens when someone creates a cute little holographic mascot thingy that sits on your desk and becomes your interactive gopher and characterful companion, but fits said cute little holographic mascot thingy with a jailbroken AI trained on horrible stuff? An AI like FreedomGPT, for example – which, in fairness, hasn’t been trained on more-or-less horrible stuff than any of the other more famous models but which instead has been tweaked so that it runs without any of the same guardrails. Want to ask it how to make bombs? It’ll tell you! Want it to give you tips on how to dispose of a dead body in such a way as to minimise detection in an urban environment? Got you covered! (An aside: to make this work properly you have to download it and run it locally – there’s a web interface, but it’s VERY SLOW). Amusingly (not hugely amusingly) this has been created as a promo by some moron VCs – WELL DONE GUYS YOU SO EDGY! To be clear, Web Curios neither endorses said VCs or any of the stuff that this bot tells you – I don’t, personally, think particular example is anything other than a silly stunt, but it’s an interesting indicator of all the fun things that are going to be coming our way in the coming months as people create their very own personalised machines – I have a slightly horrible feeling that we’re going to see a whole host of new Count Dankulas emerging, as a bunch of bedroom-based comedy edgelords compete to make their pet stochastic parrot say the worst things possible in the name of short-lived online notoriety.
- Voicechess: I promise we’re going to stop with the AI stuff shortly (no really, we are!), but before we do, another example of a GENUINELY POSITIVE USE-CASE! This is a prototypical voice-controlled chess, which is being developed by world-leading online chess platform Lichess and which you can try out yourself RIGHT NOW – all the work here is, from what I can tell, being done by off-the-shelf kit (I think they’re using Whisper for the audiorecognition, for example, though I may of course be wildly wrong here), and as such it’s a brilliant example of a useful, helpful solution to a real-life problem. Click ‘play the computer’ (or, you know, play with ACTUAL FRIENDS should you be fortunate enough to have any), click the small microphone icon in the top-right, and glory in the fact that you can now scream commands at your machine and watch as your pieces follow your commands. Because of the aforementioned natural language stuff you can get away without knowing proper chess notation, and saying stuff like ‘rook takes pawn’ actually works (although it saddens me slightly that the model doesn’t extent to ‘horsey takes bish’ and doesn’t seem to understand or enjoy it when you call them ‘prawns’), and, honestly, I think a big high-quality floorscreen on which you could play Battlechess like this, seated on thrones at either end and barking commands into the sky like some sort of mad despot king, sounds rather fun.
- Text-to-Game: Ok, this is basically just a video showing off demo technology, but I promise you that the interface on display, which seems to let you generate a navigable 3d environment from nothing but text prompts, is astonishing. Although when you stop to think about it and realise that this is going to end up being used to generate a hitherto-unimaginable quantity of CG bongo on demand it becomes…less magical, frankly.
- The Newspeak House Residency: This feels like an opportune moment to mention that my friend Ed, who runs Newspeak House in London, is currently looking for new residents to join for a year and pursue research projects around technology, government and civil society. “Newspeak House is a hub for communities working to change society with technology, spanning all kinds of civic institutions, including government, politics, activism, charities, journalism, think-tanks, NGOs, philanthropy, and academia. To quote the website, “at the heart of Newspeak House is its residential programme, running since 2015. Seven residents spend a year immersed in these communities, enjoying the chance to meet thousands of people and attend events held on their doorstep. The programme is designed to support mid-career technologists gain a holistic understanding of the civic landscape in the UK, in order to found groundbreaking new projects or seek strategic positions in key institutions. It’s ideal for people who have been working professionally for several years and are now looking to grow their network and spend time reflecting deeply on how they can best have impact on the world.” If you’re in London, or thinking of coming here, and fit the above description, and can afford to spend a year learning and thinking about your work and practice, this sounds like a genuinely interesting opportunity.
- The Undeniable Street View: This is a really smart piece of comms – The Undeniable Street View is a project by War Up Close, which is an organisation dedicated to taking panoramic views of Ukraine in the aftermath of the Russian invasion, to document the destruction being wrought on cities across the country in immersive detail, and which basically does the ‘before/after’ thing using a StreetView-style interface. You can select from various places across Ukraine, and in your browser navigate the streets of six different cities to see what they looked like before and after the war started. This is, obviously, heartbreaking, but it’s also an incredibly effective piece of comms – a combination of the first-person perspective and the familiarity of the StreetView interface make it genuinely affecting seeing the homes reduced to rubble and the potholed streets just 18 months apart.
- Visit The Global SeedVault: Virtually, that is – I mean, obviously if you want to schlep to Svalbard and see it in person then more power to you but, well, it looks cold and remote and I’m not 100% sure exactly what you’d do there once you’d got past the majesty of the architecture and the wonder of all the, er, seeds. Perhaps it’s better that you just take this virtual tour instead, and spelunk inside the remarkable project – not least because, if you ask me, it would be near-impossible not to get quite a strong wave of claustrofuturefear walking those VERY THIN-seeming corridors under all that snow (I think it’s a fact of having played too many videogames of a certain type, but it’s very hard for me not to imagine a target reticule when I first-person stroll the corridors, for example). It really is quite wondrous, when you think about it, and this is a really nice way of learning a little bit more about the Vault and how it functions and all the amazing things that they have stored in there as a hedge against the point in the not-too-distant future at which we *really* fcuk everything up for good. Seriously though, if you’re claustrophobic and the idea of exploring lots of reasonably-tight corridors underground doesn’t wholly appeal then maybe don’t fullscreen this one.
- Locals and Tourists: I rather love this – “In 2010, cartographer Erica Fischer made some simple and spectacular maps of images added to Flickr. She classified photos as either from “locals” or “tourists”, based on how far their profile location was from the photo’s geotag. These maps revealed fascinating psychogeographic patterns of urban exploration and photographic worthiness. Last summer, Dario Taraborelli suggested extending this to iNaturalist observations.” For those unaware (er, as I was til I looked it up just now), iNaturalist is a websitecommunityportalthing for people worldwide who are “naturalists, citizen scientists, and biologists built on the concept of mapping and sharing observations of biodiversity across the globe” – so basically people have given it loads of data about things that they have spotted in the wild, whilst at the same time also giving it data about themselves and where they are from. Which means, in turn, that the site is able to track which sightings of nature have been recorded by people who are local to an area, and those which have been made by people who are visiting from afar – which, in turn, means that it can offer a reasonable heuristic for tourist volumes around the world. Honestly, this is fascinating, in part to see where people go (obviously the data here skews hard towards, say, bird enthusiasts and badger botherers, but you can use it as a reasonable heuristic for general tourism imho) but also to find places to go if you’re a particular fan of arthropods or ‘herps’ (or, perhaps more usefully, should you be the sort of person who really really really prefers to avoid ‘herps’ wherever possible and wants to know where they hang out so that you can be as far away from them as possible).
- BirdBuddy LIVE: I think I featured the BirdBiddy a few months ago – it’s an internet-connected bird feeder which you can set up to snap photos of the feathery little fcuks as they dine out on your dime. Now, those of us without a BirdBuddy can live vicariously through the experience of those who do thanks to this genuinely wonderful live feed of all the images being snapped of all the birds as they happen – click the link and marvel at the steady stream of lovely birds coming at you from across the world! At the time of writing it’s all slightly-underwhelming European varietals – why are there SO MANY tits? – but if you come back later in the day you can see some truly incredible flying fauna from the Americas which I promise you are mesmerising. This is genuinely heartwarming and great.
- Keyprint: “If the internet were a country, it’d be the 7th most polluting. In the past 10 years alone, websites have grown 320% heavier. We studied the carbon footprint of the internet’s top 500 websites.” So speaks this site – unsurprisingly a marketing effort by a carbon analytics platform for websites, but one which is quite nicely-presented, and interesting in terms of how different the estimated outputs are from site to site and platform to platform. I am still waiting for someone to run the first big ‘WHAT IS THE AI REVOLUTION COSTING THE PLANET???’ server farm energy consumption scare story (it can only be a matter of days, surely?), but I was interested to see that, according to this, at least, the carbon footprint of the OpenAI domain at least is quite small – although that’s quite possibly because that doesn’t encompass the domain that they’re running public GPT on. Anyway, this is both interesting and aesthetically up my street, but see what you think.
- Japanese Woodblock Print Search: Would you like a website that lets you search an archive of over 220,000 Japanese woodblock prints, either by keyword or by uploading pictures to find similar-looking prints? OF COURSE YOU WOULD! Aside from anything else, these are just beautiful and there’s something gorgeous about the way the site lets you see the changing an evolving styles of the artists through various eras of Japanese history (there are some really gorgeous pieces of design in the latter, post-1950s period in particular). Via the wonderful Nag, this one.
- 11ftPlus8: I think every town above a certain size anywhere in the world has a BusFcukingBridge – you know, a bridge that wasn’t quite tall enough to let a full doubledecker underneath it, or a proper articulated lorry, but was both visually deceptive and appallingly-signposted and which as such each year claimed the roof of at least one out-of-town vehicle which totally misjudged the clearance and ended up shearing a few feet of metal onto unsuspected nearby pedestrians. You, er, you all had one of these, right? OF COURSE YOU DID! Anyway, this is a YouTube channel featuring footage of just such a bridge (ahem, sorry, “Railway Trestle”) in Durham, North Carolina, and if you want to see a LOT of videos of large vehicles failing to quite understand that they are taller than a bridge and, as a result, coming something of a cropper, then FILL YOUR BOOTS. A spiritual cousin of the World Bollard Association Twitter feed, there is also an accompanying website which tells me that the person behind this content has been doing this for 15 years – a degree of utterly pointless dedication that I can’t help but admire and salute.
NEXT, ENJOY SOME TECH-HOUSE COURTESY OF THIS NEW MIX BY DON’T BLINK!
THE SECTION WHICH WOULD LIKE TO TELL ANY OF YOU WHO WERE CONTEMPLATING GOING TO SEE ‘BERLUSCONI, THE MUSICAL’ IN LONDON AT THE MOMENT THAT YOU REALLY SHOULDN’T BOTHER, PT.2:
- RoboPianist: Yes, ok fine, it *is* more AI and I am sorry – BUT I PROMISE THIS IS FUN AND CUTE AND NOT SCARY! This, instead, is a rather amazing demo which accompanies a recent research paper into the development of articulated robotics; the link takes you to a bit of code that lets you watch a pair of machine-controlled virtual hands, complete with fully-articulated fingers, manipulate themselves into playing a virtual keyboard; the amazing thing here is that this is all being calculated in realtime, with the ‘fingers’ interacting with the ‘keyboard’ as though both existed in physical space, which is…slightly mind-boggling, if I’m honest. You can change the song you’re asking the machine to play, and even use your mouse to try and impede its progress – which feels mean, fine, but also allows you to see that, yes, this really is ‘working out how to ‘play’ the piano on the fly’ which, I repeat, is astonishing. I know that any sort of humanoid robot is a long way away, but the advances in articulation currently being achieved really are quite amazing.
- Chia Earth: I can’t stress enough how much I am enjoying sites like this one at present. Chia Earth is…what is it? It’s an ASCII-ish series of small meditations on Being Online. It’s a small repository of links and thinking about a different type of internet; something small and crafted. It’s, almost certainly, a bit too twee. It’s part of a growing dialogue between digital makers and creatives about what a folk internet might look and feel and sound like, and how it might work, and what it might be for. I encourage you all to have a click and an explore and a think – you can find lots of interesting things through this if you dig enough, I promise.
- Dial: You know that ever-so-slightly satisfying lockpicking minigame mechanic that you get in a certain flavour of videogame? Yes, well imagine that but significantly more inscrutable and online. As far as I can tell, this is a small lockbreaking toy which, should you manage to UNLOCK THE DIAL, will grant you an access code to shiny new alternative browser Arc – I can’t tell you whether or not this actually in fact works or whether it’s an incredibly cruel troll designed to keep new-browser-enthusiasts turning the dials for evermore as, honestly, I got bored after approximately 12 seconds, but those of you with more lockcracking nous (and, frankly, patience) than I might be able to get something moe out of this than I did.
- Snack Memories: A Twitter account dedicated to sharing pictures of old snack foods which have since been discontinued – this is runout of the US and so obviously has a North American bias, but that’s not bad thing when you consider how famously-batsh1t food across the Atlantic often is. Kellog’s Toaster Pizzas! That weird time when fast food chain Denny’s offered up “Frodo’s Pot Roast Skillet (2012-2012): Part of the themed promotional menu for “The Hobbit: An Unexpected Journey,” this meal featured slow-cooked pot roast, roasted carrots, red potatoes, onions, celery, and mushrooms, topped with shredded cheddar, served in a skillet”! The honestly repellent-sounding Nabisco Hammies, which had possibly the best slogan ever and were: “Party crackers, with the “taste of baked ham”, as well as pineapple and cloves, pressed into the vague shape of a baked ham. The slogan? “Ham tasty!…Ham shapey!””” This is genuinely brilliant and will make your Twitter experience marginally better (whilst it simultaneously continues to get worse each and everyday thanks to That Fcuking Man).
- The World Happiness Report: This made me slightly annoyed, if I’m honest, which frankly feels like something of an ironic slap in the face. Were you aware that there exists a World Happiness Report, ranking the world’s nations based on how happy they are, a quality inferred from a range of different statistical markers, each weighted to provide an overall per-capita happiness quantification? I mean, actually, now I come to think of it this *does* ring a vague bell, but the latest dataset has just been released and therefore it seems timely to link to the website accompanying it, which purports to let you explore the numbers and run country comparisons and all that sort of numberwrangling fun. You may or may not be unsurprised to know that Finland ranks top of the list (I’ll be honest with you here – the Finns I have met have not, in my experience at least, struck me as particularly full of joie de vivre, but perhaps that’s just how they react to me), with Israel. The Netherlands, Switzerland and New Zealand making up the rest of the top-10 (the UK, should you be interested, is in 27th place, which feels both…surprisingly high, and also a real indictment of all the countries below us) and Afghanistan propping up the table (you can imagine the Taliban being pretty gutted) – this SHOULD be the point where I tell you how much fun it is to explore the data and delve into the reasons why certain countries do so well, how the list changes when you look at different indicators…but I can’t say that, because the website is a fcuking broken horrorshow and the ‘data exploration’ bit is almost comically non-functional (in Chrome, at least), and…it’s just such a shame, to be honest, because this is the sort of thing which could be genuinely useful for all sorts of planning and research purposes but which is rendered entirely pointless as a result of some really crap webwork. Amusingly this is sponsored by both Walls and Unilever, who you’d think might be a bit irked at how poorly their sponcash had been spunked but who, I suspect, just signed off some monies and forgot all about why.
- Collected: This is a nice idea – collected is a site which each time you visit presents 8 different examples of good homepage/landing page design. That’s it. Each visit you get a different shuffle of the 2000+ examples it’s got in its database, meaning it’s unlikely you’ll see too much repetition unless you get REALLY obsessional, and there are some really nice bits of work in here – well done Jonas Pelzer whose work this apparently is.
- What Do People Wear In Paris?: Should anyone reading this actually know me in real life, I imagine you are sniggering slightly at the idea of me including a fashion channel thing. “But Matt!”, people who know me would probably say at this juncture, “You have worn literally exactly the same clothes for approximately three decades and they hang on you like a ‘comedy’ outfit on a lab skeleton! What do you know or care of the stylish people of the world’s most fashionable city?” To which I would probably respond “ffs must you always kick me? MUST YOU!”, but also “look, just because I dress like a blind tramp doesn’t mean that I can’t appreciate the style of other people”, and “the thing about this video series is that not only are there some pretty interesting outfits on display (if you like that sort of thing), but there’s also something really quite nice about the way in which the person filming asks questions about the lives and hopes and wants of the featured fashionistas beyond the cut of their cloth; it feels human and…nice, for want of a better word, and I found myself watching more of these than I ever expected to. Oddly enough it reminded me a bit of the OLD SCHOOL days of the Sartoralist and similar, which gave me a pleasing nostalgia-kick and reminded me of a time when someone I know was featured in it and referred to as a ‘Grup’ which I still have no fcuking clue as to the meaning of.
- Gomps: Ok, so this is QUITE technical and game development-y, but given I have no idea who any of you are and what any of you do or are interested in I suppose there’s the vanishing outside possibility that for one of you this will be aTRANSFORMATIVE piece of software. Gomps is a tool which basically lets you add multiplayer to any game built on the Unity engine – or at least very basic multiplayer features, specifically “See other players ghosts around you in real time, and leave notes throughout the game worlds for other players to read.“ Which, generally, I am a huge fan of as a mechanic – I really like the Souls-esque idea of asynchronous interplayer communications (PSEUD!) and would happily see it applied as a layer to basically every communal online experience (I still remember those nascent attempts to create a universal annotation layer over the web in the early-10s – a terrible idea, but one with SO MUCH comedic potential) – and the fact that this basically lets you do this with ANY game in your Steam library,adding a slightly social element to games that didn’t previously have them, strikes me as not only technically amazing but also a pretty cool way of exploring some emergent gameplay ideas.
- Splace: Seeing as we’re doing games, Splace is an interesting idea which is simultaneously very different from the last link and also very vaguely related – the idea of the app is that it effectively (and I appreciate I am probably simplifying the concept in ways the devs won’t necessarily enjoy so, er, sorry!) creates communities around individual games, with each title featured having its own map onto which users can add annotations, gameplay clips, tips and tricks, Easter Eggs, etc, with the idea that eventually there will be a ‘Splace’ for every single title within which its fans will be able to interact and chat and all that sort of jazz. Which, effectively, is a forum with bells and whistles, and I don’t know whether or not there’s enough here to hook people in – but, at the same time, I can see how a map-based interface collecting content relating to specific in-game locations and instances could be useful. Although, of course, it could just end up being flooded with content from moronic children – WHO KNOWS? God it’s exciting being online, isn’t it?
- MonkeyType: Look, I know that ‘learning to type better’ is noone’s idea of fun (unless you’re Mavis Beacon) but, honestly, now that we’re rapidly coming to the fag-end of the period in which ‘being marginally better at Google than your peers’ is considered any sort of competitive advantage whatsoever (I am so scared) you might want to start enhancing some of your other skills – and given we’re apparently all going to spend much of our remaining future years typing instructions to machines (until it all goes voice recognition and our fingers atrophy as our stomachs swell) you may as well take this opportunity to steal a march on your fellow prompt engineers (lol!) and learn how to communicate with The Machine marginally faster than they can. MonkeyType is a genuinely decent typing trainer, with a nice minimal style and a bunch of different training exercises that you can do to improve your performance. Look, you and I both know that this is a Cnut-like (the king, not the bowdlerised swear) attempt to stave off the inevitable obsolescence of the white collar office drone but, well, it’s worth a try, eh?
- The Blue Car: The TikTok account of a remote control car, or rather its owner – the car is used to run errands and do favours for people, and whilst you might not think that you would be amused and entertained by dashcam footage from a small RC vehicle going to the shops with a $10 bill and returning with some groceries let me assure you that you are wrong and that this is the most charming four-wheeled content you will see all day.
- Zigazoo: As the largely-confected fury about TikTok continues to swirl – gyac, everyone, if you’re worried about nefarious datacollection by bad actors then a) lol! Too late!; and b) have you checked how secure your router is?; and c) what about all the OTHER apps ffs?! – so a youth-focused competitor is making its move. Zigazoo is basically ‘TikTok for kids and with more guardrails built in’ (you will, doubtless, be heartened to learn that the platform’s ‘mission’ is to “bring out the best in humankind, build genuine connections, and cultivate a better society through positive, authentic social media built on ethical algorithms. At the largest of scales. For all”, which is nice), which promotes POSITIVITY and doesn’t allow text comments or messaging to supposedly keep it safe and clean, and the content is supposedly vetted to make sure it’s ‘positive-only’ – which, honestly, sounds MISERABLE, can you imagine a social feed filled with nothing but teenagers talking about how happy they are? Still, if you want to give a child in your life yet another reason to spend the majority of their waking hours staring slack-jawed and swiping into the eye of their portable, infinite Wunderkammer then GO FOR IT!
- Roast Me Greta: To be clear – I don’t find this clever or funny, but I do think it’s quite remarkable that someone’s been able to spin up this site which lets you click on a picture of Greta Thunberg and have a voice read out some words in her voice about what an ar$ehole you are for not doing enough to save the planet. This feels like it was created by a 14 year old – and it’s sort of amazing that it’s possible, and that the faux-Thunberg sounds as convincing as it does.
- Welcome Dream: I don’t really want to tell you anything about this. Click, explore, read, and fall down the rabbitholes of this seemingly-simple but VERY twisty and surprisingly rich little indieweb project, where every hyperlink takes you to a new fragment of…fiction? Dream? Parable? Thesis? Manifesto? I honestly couldn’t tell you, but I really really really like this. It makes me want to misuse the adjective ‘Borgesian’, which is pretty much the highest praise I can bestow.
- Sequoia Health: This made me laugh a LOT. Sequoia is apparently a ‘men’s sexual health’ app – exactly what it does is unclear, but apparently it will allow you to “Track, analyze, and improve your intimate health by executing an individually created training program developed by sexual sphere medical experts.” Yes, that’s right, SEXUAL SPHERE MEDICAL EXPERTS! There are tests you can do – beautifully, the screenshot accompanying the ‘tests’ section of the site shows the app giving a reassuring diagnosis of ‘you seem to have a healthy erection’, but HOW DOES IT KNOW?!? – and exercises it can walk you through, and a degree of what I presume is performance tracking, and, best of all, a KNOWLEDGE CENTRE featuring helpful articles and useful content, where the example they have chosen to feature is a piece headlines “Tight underwear and its impact on men’s health”. SIGN ME UP! Should you feel that this is a product you need in your life, be aware that you will need to subscribe and pay a monthly fee – but you can’t put a price on sexual health, right? RIGHT! Reassuringly the site also features a bunch of testimonials, so you can rest easy knowing that the app was awarded “Alpha Starup at Web Summit”, and is a “Partner of the Unition Nations Population Fund” and was also named as an “Impact Starup at Web Summit” (all sic), which I imagine means you’ll all be rushing to exchange your hard-earned pennies for some d1ck-diagnostics. I am sure all the 5-star reviews on the app store are DEFINITELY LEGIT.
- The Journey Planner Challenge: SO GOOD, or at least it is if you’re either a Londoner or the sort of transport nerd who knows the tube map by heart despite not being one – The Journey Planner Challenge offers you five tube stations, and you have to put them in the order that lets you travel between them in the shortest time. I promise you that this is more fun than it sounds, and EVEN BETTER it can be used as an utterly spurious “I am more London than you are” badge of honour with your girlfriend who always makes fun of you for having grown up in Swindon even though you were born in London and are DEFINITELY a Londoner whatever she thinks (what? Projecting? No).
- Dance Music WTF: The videos section of Curios is always hardest for me to write, mainly as I find attempting to describe music in prose almost impossibly difficult (or, more accurately, impossibly difficult to do without it reading really, really badly) – in part because I simply don’t really get genre minutiae and the like. Which, if this quiz is anything to go by, is because genre minutiae is absolute bunkum and means NOTHING – here the challenge is to listen to a piece of dance music and identify from a short clip whether the track in question is (for example, hardstep or gabber or nu-hardcore or technotrepanning – one of these may be of my own invention). It is in part VERY HARD, but more than that it’s genuinely amusing to finally find out what “Chemical Breaks” are meant to sound like. This feels both like something to waste time with at work and something which you could legitimately lose a whole afterparty to circa 5am.
By Useless Arm
THE CIRCUS OF TUMBLRS IS EMPTY AGAIN!
THE TROUGH OF (INSTA) FEEDS!
- Slappy The Little Green Frog: It’s a small green frog, on Insta! From what I can tell there’s noone in the comments of any of these pics screaming “THAT BEHAVIOUR IS A CLEAR SIGN OF AMPHIBIAN DISTRESS”, so I feel reasonably-safe in recommending this to you.
- Charlie Engman: I’ll be honest – I don’t really know what’s going on here, or why Engman makes images like this (unsettling ones, mainly, mixing media and human/AI created works in the feed to dizzying effect), but I very much like the style here.
LONG THINGS WHICH ARE LONG!
- The Open AI letter: You will, of course, have heard ALL about this this week if you’ve paid any attention to the news (and can I just say that I am really not enjoying all the airspace being given to people who know literally nothing about any of this stuff to opine on THE DANGERS (OR BENEFITS) OF AI – I am looking specifically at ‘lazy person’s go-to intellectual Yuval Noah Harari, whose interview scaremongering about AGI on Radio4 this week practically had steam coming out of my ears, so unhelpful and unanchored in practical reality was it), but if you’ve not read it then it’s worth taking a look at the actual document to see what Gary Marcus, Elon Musk et al have actually signed and what exactly they are calling for when they ask for a ‘pause’ on the development of post-GPT4 iterations of LLMs. There are LOTS OF OPINIONS about this everywhere, and, honestly, you don’t really need to hear mine beyond my general feeling that LOL ONCE AGAIN WE WORRY AT THE BOLTS OF THE GATE AND THE OILING OF THE HINGES WHILE THE HORSE CAVORTS HAPPILY IN DISTANT MEADOWS, but, should you be interested in a counterargument, this piece does a reasonable job of articulating one or two objection to the position outlined in the letter (mainly: that the dangers are already here and that worrying about ‘the future’ rather neglects to pay attention to the present; and that we might want to think a bit harder about all the things that are going to happen as a result of the fact that the current iteration of this tech is already here and already in the wild). I agree that this stuff needs more thought, and that it needs regulating, and that ploughing headlong into implementing untested technologies whose impacts we don’t necessarily understand is A Bad And Silly Thing To Do – but, look, that’s already happened, so can we maybe instead focus more energy on thinking about what some sort of universal regulatory framework might look like, and how one might go about creating enforceable standards, and all that really boring stuff that actually matters (on which point, if you’re interested in this then it’s worth reading the UK government’s consultation paper on AI which was published this week – it’s quite wooly to be honest, and there’s an awful lot of handwavey ‘the regulators will sort this out’ chat which will be…amusing to anyone familiar with the effectiveness of Ofcom and the rest, but it’s a useful example of how state actors are currently thinking about AI and its implementation).
- AI Calluses: I’ve not touched on PuffaPope in here because it’s been discussed ad nauseam elsewhere (although, briefly – the genius of the image imho was that its plausibility lay in the fact that it acknowledged that as an adopted Italian it is OBVIOUS that the Pope would totally rock an outsize Moncler), but it feels like a bit of a rubicon in terms of the AI pollution of the information wellstream; this piece by Dave Karpf isn’t about the Pope, but does touch on similar territory. In it, he writes about being questioned by a student about a particular paper he’d previously written – except, when he went back to check his records, he realised that he had never in fact written the work that the student was citing, because it had been made up by GPT as part of its answer to one of the student’s question. Karpf frames this as the start of the introduction of a new layer of social/digital friction, whereby we are all just going to have to get used to the necessity of an extra layer of factchecking in all of ouf digital interactions from hereon in – because, honestly, you really can’t take ANYTHING at face value anymore.
- Fake History: A thematically-adjacent piece in VICE, this article looks at the boom in people using the latest version of Midjoiurney to imagine alternate moments in history; now that anyone can ask The Machine to generate a near-photorealistic depiction of, say, ‘The Great London Frog Rain of 1911’, what do we do about the danger that these false narratives and made-up events get eaten and chewed abnd digested by culture and believed? This follows neatly on from that essay a few weeks’ back about the importance of photography as reportage, and the danger of what happens when we cease to be able to take ‘photojournalism’ at face value – look, I am a desperately cynical person who believes the worst of everything and everyone (it’s a fun time, being me!) and even I am preemptively-exhausted at the idea that everything has to be questioned. Does anyone want to have a meaningful conversation about the renewed importance of teaching critical thinking at a young age? Eh? Oh.
- The AI Election: I’ve worked in and around politics and campaigning enough to recall three separate instances where we were told we were experiencing “THE FIRST SOCIAL MEDIA ELECTION” – we can look forward to the (already horrifying) prospect of the US Presidentials in 2024 being dubbed “THE FIRST AI ELECTION”, and, yes, that’s likely to be exactly as fun as you can imagine. This is a NYT piece which takes a relatively sober view of all the ways in which the current crop of AI tools (which, let’s all remember, will have been surpassed in ways we can’t quite conceive of by the time the States gets to the business end of the whole sorry saga in 15m time) might be used as part of the campaigning process by lobbyists and marketing campaigns, from automated targeted emails to the more sinister end of the spectrum (faked audio used in hypertargeted attack ad campaigns, to name but one example). Again, the sort of thing that makes you realise that worrying about ‘the future’ of this stuff is probably less helpful than worrying about its very real present.
- AI and the American Smile: A really interesting article which makes the smart observation that all the ‘selfie’ style of AI-generated photography (which leads to stuff like “A selfie of roman centurions outside the colosseum”, say, or “A selfie of hitler and the lads in the bunker”) has been trained on a corpus of images which is largely North American or European, and which come from cultures where the role of the smile in interpersonal relations is specific and distinct from the way in which people present in, for example, Slavic or Middle Eastern cultures. Which, fine, is a small thing, but once again it links to the ways in which we are (sorry for the hyperbole, but) rather polluting the truthwell with this stuff in ways whose impact we really won’t understand for ages.
- Bots Write Jokes: A little post by Rob Manuel explaining how he used GPT to write jokes about humans from the point of view of robots – I always enjoy people explaing their working, and this is a nice, short post which talks you through how Rob got the prompts to work and which is a useful set of guiding principles for your own interactions with this sort of software. Also, as Rob points out, “it’s more fun to play with this technology than it is to sit on the sidelines worrying that everything is fcuked. Have fun, life is short etc” – which, obviously, isn’t advice that I personally take but which I heartily advise YOU to.
- The Substack Question: You may have noticed this week that a bunch of Substack writers sent out posts telling you how excited they were to be, er, investing in Substack. Presuming that you didn’t read those – because, honestly, why do I fcuking care that you’re investing in the platform you publish on, you weirdo? What do you expect me to do ffs, clap? – you might be interested in this article in The Verge in which Elizabeth Lopatto looks at the underlying economics behind this push and what it tells you about Substack’s revenue and its overall business. Ok, fine, this isn’t super-compelling (though I personally enjoyed the tone), but it’s a really good piece for anyone interested in the economics of tHe CrEaToR eCoNoMy et al.
- Fortnite and the Metaverse: LOL THE ‘M’ WORD LOL! Do you remember last year when all those people were trying to sell you b2b metaverse solutions? WHAT A TIME TO BE ALIVE (and, er, while we’re here, I’d probably hold off on buying any b2b AI solutions from those selfsame people who have seemingly all pivoted hard in the past 12 months)! Still, despite ‘the metaverse’ continuing to be at best a nebulous and poorly-defined concept whose closest actual definition might as well be ‘videogames’, there is still a significant degree of interest in the idea of interoperability between virtual spaces, and a significant number of people socialise in embodied virtual space every day – and this interview with Epic Games CEO Tim Sweeney (Epic’s the company behind Fortnite), in which he talks about the platform’s recent moves to share revenue with users who create games and experiences on it and the way in which he envisages these digital spaces evolving, is genuinely exciting and made me once again think that there’s something in this space which could develop into something really interesting as soon as Zuckerberg fcuks off and leaves it alone. BONUS VIDEOGAMEMETAVERSE CONTENT: this is an interesting look at how Roblox (the only other company with a realistic claim to be doing anything meaningful in this space right now) is integrating generative AI into its creative tools for players and developers.
- Care Robots: A New York Times article looking at Italy’s care crisis and nascent plans to try and introduce robotic assistants to the care sector in a bid it alleviate the burden on families and careworkers from Italy’s ageing (and, honestly, seemingly never-dying – NOBODY NEEDS TO LIVE THAT LONG, ITALIANS! TAPPING OUT IN YOUR 80s IS LITERALLY FINE!) population. As someone with first-hand recent experience of caring for both a geriatric invalid and a terminally ill person in Rome, I can testify to the fact that there is no way that the Italian care system is equipped to cope with another decade or two of growing numbers of old people whose families can no longer afford to take care of them at home and where the illegal foreign labour they’re relied on is being chased out of the country by a racist government, Bring on the robots, basically.
- Bicycle: The latest in Bartosz Ciechanowski’s ongoing series of interactive explainers, this is a genuinely brilliant (long) read explaining exactly how a bicycle works – honestly, I don’t think I’ve ever seen someone who is so GOOD at explaining hard concepts using small interactive examples, and I say this as someone who not only can’t draw a functioning bicycle but who is so bad at physics and mechanics that they couldn’t even explain to you WHY their ham-fisted attempted drawing of a velocipede wouldn’t in fact work in real life. This is superb, and I really hope that Mr Ciechanowski parlays all the attention that his work is getting him into some suitable work or remuneration because he really is world-class at this.
- Ghost Ships: I concede that you might not automatically think that reading several thousand words about the potential automation of the global shipping trade is a good way to spend ten minutes of your life – and, you know what, maybe you’re right. Still, I promise you that this is LOADS more interesting than it has any right to be, and asks really interesting questions about what the potential impact of a near-faultless realtime vision of all of the shipping data, wrangled by AI to optimise cargo and routes and associated things like prices and supply and advertising, might be. Honestly, this is one of those articles that starts relatively small but which by the end has you reeling at the complex interconnectedness of everything in modernity (or at least it does if you’re a dullard like me).
- Dumb Turning Points in History: Ok, fine, this isn’t an article so much as a list of Tweets, but some of these are SO GOOD. My personal favourites in the canon of this sort of thing are: a) the fall of the Berlin Wall happening when it did because the guy doing the press conference hadn’t been given a proper Q&A briefing, and as such when the first question from the press was the inevitable “so, what’s the timescale for the unification of East and West Berlin, then?” the spokesperson floundered and just sort of said “Er, now?”, at which point history started happening; and b) the fact that Macau was owned by the Portuguese for 400+ years as a result of them being the principle traders of ambergris in the 16thC, and ambergris at the time being a favoured aphrodisiac used in the preparation of tonics for male, er, ‘vigour’, and the Chinese emperor of the age having one or two problems enjoying the 100 virgins he’d been gifted by some potentate or another, and as such the Chinese swapping an entire island colony for balene Viagra because their head of state was an impotent, middle-aged man (this anecdote, by the way, from Lucy Inglis’ fantastic book about the history of opium).
- Visiting 90sCon: Rolling Stone takes a trip to a 90s convention, where fans of all things, er, 90s, can queue up to see panels featuring former members of N*Sync (have I punctuated that correctly?), or to spend $50 to get Shannen Doherty’s autograph (one the one hand, if someone offered me $50 for my signature I would bite their fcuking hand off – please contact me via the usual channels should you be in the market for some genuine Muir scrawl!; on the other, I don’t think I can begin to imagine the weirdness of the experience of sitting for an hour while people file past you watching you write your own name) – this is a gentle read that will give you some not-insignificant nostalgia flashbacks if you’re anywhere near as Methuselan as I am.
- Azealia Banks: It’s something of a shame that the artist responsible for one of the most jaw-dropping breakthrough singles of the century so far – it still bangs – has become better known for being a, ahem, ‘loose canon’ than for their music; still, this profile of Banks in Dazed seeks to present her in a slightly more music-and-art-first light than has been the case in recent years, and does a pretty good job of explaining (if not necessarily justifying) some of her more…extra moves. There are some good points in here, not least the perceived difference in treatment between Banks and Tyler the Creator and other (male) artists who’ve said/done one or two…controversial things over the years but who seemed to have been rehabilitated just fine thankyouverymuchindeed, but personally I feel the article feels a bit too much like it was proofed by Banks herself to ensure the requisite degree of ‘godlike genius hagiography’ which is sprinkled (to my mind unwarrantedly) throughout. See what you think – if nothing else, it’s an interesting read.
- Brandon Sanderson: I hadn’t heard of Brandon Sanderson before reading this article – or rather, I probably had done at some point, but only in a very abstract way (“This obscure fantasy author is THE most successful crowdfunder ever” sort of thing) – and, unless your a fan of a particular brand of epic fantasy novel, it’s unlikely you will have done either. Despite his lack of fame, though, Sanderson is a notable writer, in terms of the prolific quality of his output (seriously, we’re talking MILLIONS of words here – from one logorrhoeic to another, I salute you Brandon) and the devotion said work inspires in fans worldwide. Which combination of lucrative-but-niche fame and relative obscurity (and his Mormonism, which journalists will never cease to enjoy writing about) probably led WIRED to run this interview, in which one Jason Kehe goes and hangs out with Sanderson and his family and…and…well, is just spectacularly fcuking mean about the man, to be honest, criticising his writing and his personality and his fans and the genre he writes in and his ideas and, honestly, this is SUCH an incredible and weird hatchet job (which, at the same time, is also a readable article, which makes it…worse, somehow) which is made all the more stark by the INSANELY gracious response Sanderson went and posted on Reddit and which made me think that however awful his novels might be, the man himself is probably quite a nice guy (please, noone milkshake duck him quite yet).
- The Story of a Picture: A wonderful essay, this, from the Economist, looking at a truly iconic (sorry) photograph from pre-war Britain which depicted (or so viewers were led to believe) the class extremes that bookended the nation. ‘Toffs and Toughs’, as the picture came to be colloquially known, showed a bunch of Etonian boys in one half of the pic contrasted with a bunch of ‘street urchins’ on the other side, and became something of a recurring theme whenever anyone wanted to illustrate the English class divide. Except, well, the actual story behind the picture isn’t quite that simple. This is SO interesting, and a sort of object-lesson in sources and authenticity and fact-checking and how stories are more powerful and the truth, and, weirdly, it’s an excellent companion to those articles at the top of this section about AI and imagery and authenticity.
- How Not To Dig Your Own Grave: Claire O’Brien writes about death and the business of death, about her experience working with dead bodies and what you learn about yourself and other people when you’re hanging out with cadevers on the reg.
- Doggerland: Finally this week, an excellent piece of short fiction by Kaliane Bradley which this week was announced as the winner of the VS Pritchett Short Story Prize. This is GREAT – funny and sinister and atmospheric and classical and modern, and it contains some truly wonderful lines, and it’s an excellent piece to lose yourself in of an afternoon.
AND NOW, MOVING PICTURES AND SOUNDS!: