HELLO EVERYONE HELLO!
Well, we have a new king, but more importantly we have NEW KNOWLEDGE – specifically, we learned that there is literally NOTHING that the UK is hornier for, sexually or electorally, than a middle-aged woman carrying a ceremonial sword. Were it possible for an entire nation to undergo psychotherapy I would be strongly advocating for it right now.
I have errands to run, and you have an absolute FCUKTONNE of words and links to click through (you…you will click, right?), so let’s get started shall we (Jesus, I appear to have taken to writing this upfront in the sort of stern, authoritarian, vaguely-schoolmistressy tone that I imagine Penny Mordaunt to speak in – I can only imagine the tumescence this is causing)?
I am still Matt, this is still Web Curios, and you don’t have to be embarrassed, we’re all friends here.
PS – there’s a chance there won’t be a Curios next week, so on the offchance that there isn’t then don’t worry, I am probably not dead. Although I might be. Consider me in an exciting state of Schrodinger-ian uncertainty between now and the next edition!
By Ralph Gibson
THE SECTION WHICH BELIEVES THAT THERE IS ONLY ONE NEWS STORY THAT COUNTS THIS WEEK AND THAT THIS IS IT, PT.1:
- The Stable Diffusion Photobooth: I promise that this week’s issue is comparatively light on AI stuff, but I couldn’t help but lead with this because, honestly, it is SO COOL. Oh, ok, fine, it’s far too steampunky in design to be ‘cool’ in any meaningful sense (I am sorry, but steampunk is the opposite of ‘cool’ – I don’t make the rules, it just is), but it’s SO MUCH FUN, and exactly the sort of genuinely-creative and slightly miraculous use of AI that I personally would like to see more of (and which therefore is almost certainly doomed to obscurity). The link takes you to a video on Reddit which demonstrates the user’s ‘intern project’ (genuinely curious, parenthetically, as to what sort of internship they are undertaking), which is basically a ‘photobooth’ that they have hacked together out of a bunch of old kit including a rotary phone and a greenscreen monitor – with a bit of analogue fiddling, including flipping switches and turning dials that seemingly correspond to settings for the AI, the machine eventually prints out a Stable Diffusion-juiced portrait of the user, just like an actual photobooth but FUN and UNPREDICTABLE and VAGUELY MYSTERIOUS. I appreciate that even by my low standards this is a particularly-mangled car crash of a description, so please click the link and see it for yourself and then imagine how many other EXCITING AND DELIGHTFUL physical installations you might create by mashing together all this fun new tech. Honestly, if nothing else I would LOVE someone to make one of those fortune telling ‘end of the pier’ machines (as popularised in Big, for example) cobbled together from a GPT and some avatar-generation software. In fact, now I come to think of it what I would really like is a whole modern-but-retro arcade full of AI-enabled toys like this, so if one of you could possibly fcuk off and make that a reality for me that would be great thanks.
- Peridot: On the one hand I know that by linking to this I am in fact simply doing the marketing for a massive company; on the other, IT’S BASICALLY AR TAMAGOTCHI! Peridot is the latest attempt by Niantic to recapture the lightning-in-a-bottle success of Pokemon Go, except this time with its own brand new IP – Peridot, which I must stress I haven’t been able to try out due to the fact that either my phone is simply too sh1t or it’s not enabled in the UK yet, is a game/toy/thing which lets users create and care for their very own imaginary digital creatures (the ‘Peridots’ of the title), which they experience through AR and which can be fed and interacted with and played with and groomed and, most excitingly of all (if what gets you excited is imaginary creature eugenics), BRED! Yes, that’s right – as you can see in this gameplay preview video (skip to about 10m in if you want the HOT BREEDING ACTION), Peridot owners can if they desire choose to ‘mate’ (yes, ok, fine, Niantic doesn’t use this terminology, but let’s just accept that the Peridots fcuk and move on) with another user’s creature to create a brand new…thing, with characteristics drawn from each parent (beautifully, as evidenced in the gameplay vid, this is occasioned by users seemingly ‘showcasing’ their pets in specific locations which marks them as being available for breeding which, look, I can’t help but admit seemed a bit…well, pimp-like, if you ask me). This only launched this week and as such it’s still quite light on practical detail as to how it all works – one thing that is clear, though, is that it contains a fcuktonne of microtransactions because it’s 2023 and that is how this stuff works these days (slightly miserably, it’s also being used to showcase a new Amazon feature whereby you can link your Amazon account to the game and, inexplicably, order a Peridot sweatshirt from within the app – why? WHY THE FCUK NOT?), so I would advise anyone thinking of letting their kids play with this to be aware that they may be opening some sort of infinite money pit by so doing. I am genuinely curious as to whether this will take off – regular readers will know that I have been banging a tedious drum about the possibility of AI-enabled digital pets as a ‘thing’ for a while now, and this feels like the first toy to really explore that possibility in a meaningful fashion.
- The GigaBrain: This is genuinely brilliant. You know how for a few years now Google has been increasingly useless, and people who know have been basically just using it as a way of searching Reddit for specific advice or product recommendations? Well, the Gigabrain is a search engine which basically removes the tedious hassle of having to include ‘site:Reddit.com’ at the start of your search – it’s a, er, search engine which only looks at Reddit, and uses some light natural language processing to interpret questions and deliver results and, honestly, from the little I’ve used it this week it seems genuinely pretty good; give it a go (oh, and seeing as we’re ‘doing’ search, here’s a rundown of all the Google announcements from this week which I honestly can’t get excited about because I am fundamentally dead inside (and also because I long ago stopped believing that iterative technological improvements would in any way ameliorate my life) but which you might it find useful to be aware of).
- The Bluesky Firehose: Are you on Bluesky? Are you? WHY HAVEN’T YOU INVITED ME THEN?!?! Ahem. To be clear, I have limited interest in being on Bluesky and so you can keep your invites thankyouverymuchindeed, but if YOU are desirous to be in with the ‘cool’ kids (I’ll leave you to determine the exact ‘cool’ value of achieving relatively early access to a text-based online messaging network) then why not press your nose up against the glass and enjoy this peek behind the velvet rope. This site is basically a rolling feed of everything posted on the platform, which is still vaguely parsable given it’s still in beta and pretty unpopulated – so what sort of posts make up the Bluesky ecosystem? To be honest it’s hard to get any real sense for the platform’s vibe from this feed – it’s completely acontextual, and there’s no sense from the way the information is presented of who is interacting with whom in any given post, and as such there’s actually something vaguely soothing and poetry-like about the decontextualised words flowing across your screen. That said, I think – based on the chat visible at 735am this morning, it’s fair to say that it’s your standard mixture of single-issue lunatics ranting into the ether, sh1tposters and, from what I can tell, a surprising number of people who really, really want to talk about Digimon. Does that sound like a party you want to be a part of? WELL WAIT YOUR TURN. BONUS ADDITIONAL SOCIAL NETWORK INSIGHTS: this site collects the most popular posts and links across the wider Mastodon Fediverse, should you wish to be able to take a temperature reading of what those people are currently obsessed with (I recommend looking at the ‘Posts’ tab, where right now there are a lot of photos of pretty good animals).
- Caryn: This is interesting, and, quite possibly, the first in a coming wave of similar projects. Caryn – and you’ll have to bear in mind that everything I am about to tell you is gleaned from research and speculation rather than first-hand experience because there is no fcuking way I am paying money to actually try this, sorry – is a project by a real-life Snapchat influencer who has basically made an AI persona out of themselves and is charging for access to the model; users can sign up via a Telegram channel to connect with the bot, with interactions priced at $1 a minute. There’s an article about the whole thing which suggests that in 10 days since launch the real-life Caryn has coined over $70k from (presumably) horny teenage boys, which is…staggering, frankly – if you’d like to see what exactly they are paying for you can watch a short video of someone interacting with digital Caryn here, although I’m willing to bet that the vanilla chat they showcase is not necessarily representative of the majority of interactions with the model. It’s voice-and-text only, with no video at present (although I’d imagine that that’s an inevitable future build), and, honestly, leaving aside the general ickiness of creating a digital puppet of yourself that you know is going to be used as masturbatory fodder by online strangers of questionable provenance, it feels like this is very much going to be A Thing for famouses in the short-to-medium term. Just you wait – it surely can only be a matter of time before you’re able to have an officially-sanctioned AI version of, say, Richard Madeley on your phone to do whatever sexy bidding you demand of it (I don’t know why Richard Madeley sprung to mind; possibly because of this).
- Explore Mars: Would you like to experience what is basically Google Earth but, well, for Mars? OF COURSE YOU WOULD! And now, thanks to the magic of this website, you can! Pretend you’re the Mars Rover, all cold and alone and distant, pootling about the surface of the red planet with not a care in the world! This is quite amazing – compiled from photographs and data captured by a succession of probes and drones, including Spirit and Curiosity, I don’t think we’re ever going to be quite as amazed as we ought to be by the frankly mind-flaying fact that we’re able to sit here and explore a literal other planet while sitting at a desk and having a cup of tea. There is SO MUCH to explore here – the video ‘flythroughs’ in particular are quite amazing – so do have a click and a scroll and a wonder and a gawp.
- Patterns: I really don’t understand what’s going on on this website or quite how it works or what it’s for (this is the high-quality analysis and insight that you all subscribe for, right?) – look, YOU click the link and take a look and try and work it out. See? Fcuking impossible. All I know and all I can tell you is that you can use it to make pleasing and slightly-organic-feeling little line pattern drawings which are oddly-beautiful and whose animations as they grow are weirdly…biological – I can best describe this as ‘the sort of visual toy that I can totally imagine staring at for a full ten minutes, slack-jawed, were I very, very stoned indeed’. Which I’m sure is massively illuminating for you – no, really, you’re welcome!
- Soundwaves World: OH LOOK IT’S A METAVERSE! I’ve stopped including much of this stuff because, well, it feels a bit like kicking puppies or shooting fish in a barrel at this point (I do wonder, briefly, what the conversations are like between agencies and those clients who they last year managed to persuade to spend six figures on some p1ss-poor 3d world – does…does anyone mention it? Or does everyone pretend that it didn’t happen, much like the NFT drops in 2020?), but this is marginally-less pointless than almost every other version of these things I’ve seen in the past few years and you may enjoy it. I have literally no idea who made this or what it’s promoting (sorry, but I don’t care), but it was seemingly created to tie in with this year’s SXSW and is an archive of a bunch of different performances by artists (none of whom I have ever heard of, but that’s probably due to me being old rather than said artists being nobodies) which loop on the hour and which you can experience IN THE METAVERSE! So you get the standard ‘WASD your avatar around a rather underwhelming 3d environment’ schtick, but with the added ability to see vaguely-photorealistic 3d representations of a selection of artists (you can choose which performance to see in the settings menu) capering gigantically around you, in the manner of the now-classic ‘gig in Fortnite’ activation. This is, let’s be clear, still a crap way of ‘experiencing’ anything, but I didn’t totally hate it and at least there’s a reason for users to hang out in this otherwise-empty digital playground, which is an improvement on literally 99.9% of this stuff.
- RTRO: A new social network! How exciting! RTRO, as the name suggests, wants to take you back to the GOOD OLD DAYS (no, really, this DID exist!) of social media, with REAL PEOPLE and GOOD VIBES and NO ALGORITHMzzzzzzzzzzzzzzzzz…oh, sorry, it’s just that I am so fcuking bored of people attempting to recapture the never-to-be-felt-again novelty and possibility of being connected to the rest of the species for the first time ever. YOU WILL NEVER FEEL LIKE THAT AGAIN IT IS LIKE YOUR FIRST KISS IT IS GONE. Still, RTRO has a few interesting gimmicks – the split in feed between a ‘brands and commerce’ side and a ‘real people’ side is interesting in theory, although I think the fact that noone in their right mind would ever choose to click on the ‘brand’ side of the feed does rather limit the potential for this to monetise successfully, and the idea of AI prompts to cajole people into posting feels like a grim vision into the imminent future of engagement-juicing hacks – and who knows? If you’re after a ‘positivity-focused social platform that takes you back to a time when social was simply about real content and meaningful connections’ (their words, which, to be clear, I hate) then you might enjoy playing with this.
- The Coney Zoom Bar: Do you remember 2020, and the early days of the pandemic, and that weird period where we all experimented with doing theatre and comedy and quizzes on Zoom? Well in case you feel vaguely-nostalgic for that period (you weirdo) you might enjoy this selection of interactive theatrical events being put on over the summer by excellent and playful makers of EXPERIENCES, Coney. I haven’t personally done any of the shows listed here, but I’ve enjoyed much of Coney’s work in the past and in general can highly recommend their work – if you’re in the market for some gently-playful internet game theatre fun then you will very much enjoy this stuff (oh, and Coney were royally-fcuked by the Arts Council in the last round of funding, so this is also an opportunity to help support the arts at a grassroots level should you care about that sort of thing).
- Reclaim Your DNA: As far as I can tell, most booze marketing seems to involve attempting to associate your brandw with FUN without at any point being seen to be making the entirely-spurious claim that ‘being drunk’ is in any way, you know, enjoyable – so MILLIONS OF POINTS to this campaign by Nigerian stout brand Trophy, which is entirely focused on getting back all the historical Nigerian artefacts that have been looted by European colonials over the years (er, that’ll mainly be the Brits then. Again). There’s a petition to sign, and you can download AR filters for Insta that are based on the masks, and 3d models of the specific artefacts to use how you see fit, and basically it’s just really cool to see a drinks campaign that is doing more than vaguely alluding to music and sex.
- Diem: This is a really interesting idea that I can’t see working at all, but the combination of crowdsourced wisdom and an AI interface seems like possibly fertile territory. Diem is a female-centric social network and knowledge platform which is designed to turn the conversations and oral knowledge base of women’s communities into a useful and searchable resource for women worldwide. Says the blurb: “We’ve trained a model called Diem AI to conversationally answer your pressing, personal, embarrassing, funny, and serious questions. The model combines an LLM with our own data model (meaning the knowledge shared by community members in our platform). When you start a Diem (aka. ask Diem a question), you’ll receive an AI-generated response that scrapes Diem (and the internet) for answers, through a feminine lens, and then supplements those results with real-life anecdotes shared in Diem. Think of it as a Q&A sesh that’s similar to searching the web, but with a built-in network of trustworthy internet friends. Right now, our community has mostly been sharing stories about personal health, money, and relationships. They’re ****all pretty taboo topics, and that’s the point. You can also “contribute” to a Diem by sharing your own stories and recommendations, either in voice note form (a new feature!) or by writing it out.” Now obviously the main issue with this is the size of the corpus – it’s not useful without a lot of material to draw on, and getting that material is HARD – and it’s that that makes me think that Diem is sadly doomed to obscurity, but I admire the ambition on display here.
- To Be Build: Pictures of buildings and construction! “A visual journey into how buildings come to be. Celebrating the creation of architecture, the beauty of building sites and the indispensable labour involved in the process.” I really, really love this collection – there are SO MANY great pieces of photography here, whether or not you’re particularly interested in seeing pictures of girders and cranes.
- Postcards from Timbuktu: I received a nice email from a guy called Phil Paoletta this week whose email signature featured this url – WHAT A BRILLIANT SERVICE! Do YOU want to receive a postcard from Timbuktu? OF COURSE YOU DO! This is so so cool – get yourself a postcard sent from Mali and give some work to unemployed former tourist guides whilst so doing. Even better, use this as the basis for a long-running and potentially psychologically-ruinous prank on a colleague or family member – the postcards can be inscribed with whatever text you desire (although I presume Phil baulks at stuff like ransom notes or threatening demands for low-denomination currency in vast volumes) so you really could use this to fcuk with someone’s head on a vast and terrifying scale should you choose.
THE SECTION WHICH BELIEVES THAT THERE IS ONLY ONE NEWS STORY THAT COUNTS THIS WEEK AND THAT THIS IS IT, PT.2:
- Raybot: The news that Achewood was back with all new content sent certain sections of the web into happy paroxysms – there was also some fear that Chris Onstad was using AI to help him write the new strips, but instead Onstad is doing something far more interesting (to my mind, at least) by feeding his writings to The Machine and using the corpus to create an interactive digital avatar for Ray, one of the main Achewood characters. You can interact with Ray at the main link here, asking him questions about anything you like to which he will respond in pretty decent Ray-ish prose (if you know Achewood then this will obviously work better for you – if you don’t, well, MORE FOOL YOU); I am quietly excited about the possibilities of stuff like this as we move towards the ‘everyone has their own personal AI that they have trained to be whatever they need it to’ inevitable future.
- Narrative Debris: OH GOD I LOVE THIS. A gorgeous artwork exploring a very specific place, Narrative Debris is…oh, look, let me just use Tricia Enns own words here, as they will be better than mine. “This work currently revolves around the Quartier des Spectacles neighbourhood in Montreal. A neighbourhood close to the St. Laurence River and thus the arrival point to the land and later the city for many groups of people such as: Iroquoians, French explorers, Chinese, Jewish and Portuguese immigrants to name a few1. The area continues, to this day, to be a tapestry of stories, but that plurality has been threatened by recent development. How do we share, propagate, and celebrate the many stories held within such a small area?” Through a guided walk available on Soundcloud which she created but anyone can take, and a hand-drawn map of the area she’s detailing which is marked and annotated with the things she and others have seen and experienced on said walk, Emms is building a picture of a neighbourhood from sounds and memories and experiences, and this site is a tiny digital diorama of the people and memories that make up the place. Honestly, this is practically-perfect in every way and I adore it immoderately.
- Bring Back YouTube Annotations: OLD internet people like me (and, quite possibly, like you – WHO ARE YOU?) will remember the glorious (oh, ok, that’s hyperbolic; it was of course merely FINE) era of YouTube annotations which allowed creative an enterprising people to create all sorts of surprisingly-deep interactive experiences by cobbling together videos through embedded annotations and hyperlinks – YouTube killed that a few years ago, thereby sadly breaking all those projects, but now there’s this Chrome extension which will apparently restore the functionality and let you once again enjoy all the ‘Choose Your Own Adventure’-style videos that you remember so well from that brief period circa 2011 when they were very much in vogue.
- Star Trek Bridges: Do YOU like Star Trek? Do YOU secretly make the small ‘whoosh’ noise that the doors in the TV show make whenever you approach an automatic doorway? Did you harbor illicit erotic thoughts about Spock/Sulu/Uhura/Riker/Troy/Worf? Do you know the difference between a ‘Trekkie’ and a ‘Trekker’ (apparently it’s something to do with whether you can actually speak Klingon)? If the answer to any of these questions is ‘yes’, then this website will probably render you almost painfully-tumescent – it lets you ‘explore’ CG representations of the Bridge of a whole bunch of different spaceships from the Star Trek universe (don’t ask me what they are, I am neither a Trekkie nor a Trekker) via a really, really horrible web interface which does an excellent job of aping the horrible UI you’d imagine a starship operating system from the 1960s would have.
- Squirrelcam: Technically this is a camera in a bird’s nest, but for some reason the birds have been evicted by some squirrels which have just given birth in there – apparently squirrels spend about 10 weeks suckling their young, so bookmark this and check in daily for the next couple of months to see lovely, cute tree rats growing and thriving. Beautifully there is a chat window on the site which seems to be full of Belgian rodent enthusiasts all typing “I LOVE SQUIRRELS!” which is, honestly, very cute.
- This Website Will Self Destruct: One of two websites this week which I am re-featuring (breaking a self-imposed rule – THE STANDARDS ARE SLIPPING!) – This Website Will Self Desctruct celebrated its third birthday this week, which is slightly astonishing considering its whole thing is that it will disappear if noone submits a message via the contact form for 24h. What that means, of course, is that there are hundreds if not thousands of anonymous messages from strangers buried in the site’s backend, and you can explore them by clicking ‘Read a message’ at the bottom of the page. Honestly, I lost about 20 minutes to this earlier in the week – I am a total sucker for anonymous memoryholes like this, and there is something almost infinitely poignant about seeing all the people who express their own feelings of sadness and loneliness and happiness and fear via the medium of messages that will in almost all cases never be read by anyone, ever. I know that I am a bit fixated on this stuff, but I would honestly sit in a gallery watching these come up forever.
- Listen To Wikipedia: Our second reappearing link of the week is this one – PROPERLY old, this, maybe even ten years, but still wonderful and worth re-upping because I think they recently updated the code to improve the sounds and include edits from Wikis in non-English languages, making the pleasing plinky cacoophony less sparse than it previously was. Oh, hang on, I haven’t explained what this actually is – it makes a sound every time an edit is made to Wikipedia. There.
- Goblin Tools: This is an interesting idea, and if you are (or know someone who is) neurodivergent then you may find it a useful collection of helpful tools. “goblin.tools is a collection of small, simple, single-task tools, mostly designed to help neurodivergent people with tasks they find overwhelming or difficult. Most tools will use AI technologies in the back-end to achieve their goals. Currently this includes OpenAI’s models. As the tools and backend improve, the intent is to move to an open source alternative.” So there’s one that checks copy for tonal implications that might not be obvious to people with certain conditions, for example, and another which tells you how long you might usually expect a certain task to take – this strikes me as a smart use of OpenAI tech and, in general, A Good Thing, and the sort of thing which could usefully be worked into other sites and services without too much hassle.
- The Index: This is interesting: “The Index is a curated online gallery with the best design studios, designers, type foundries, and other creatives worldwide. We aim to publish a handful of submissions per week, both on the website and featured in the weekly newsletter. We also post and promote on Instagram and Twitter. We strongly believe in quality over quantity, hence our relatively slow growth approach.” There’s something perhaps *slightly* grifty about their insistence on charging $25 per submission – with no guarantee that submissions will make it onto the list, of course – but I equally admire the bloody-mindedness of the enterprise. There are currently 48 studios on there, with a reasonable geographic spread, so should you be looking for vetted design partners then this is worth bookmarking I think.
- Condiment Packets: Have you ever held a small sachet of ketchup or mayonnaise in your hand and marveled at the elegance and simplicity of its design and the almost supernatural allure of the branding and logo work? No, of course you haven’t, you’re not a mad obsessive with an inexplicable fetish for the world’s most-polluting byproduct of late-period capitalism (or at least I presume you’re not; as previously discussed, though, I have no fcuking idea who you are. WHO ARE YOU?). That said, SOMEONE is definitely that sort of mad obsessive, or at least I presume they are given the obvious time and effort that goes into maintaining The Condiment Packet Gallery, which has, according to its ‘About’ page, been going for 20 years. Annoyingly I found myself getting weirdly into this, to the point where I found myself looking up exactly what is in the mysterious ‘Salsa Golf’ marketed by Hellman’s in Argentina, so don’t for a second presume you’ll be immune to the mysterious allure of the condiment collection.
- Theoretical Puppets: I am slightly upset that I don’t appear to have featured this already in here, as it is SO GOOD. Have you ever wanted a YouTube channel that explains some pretty knotty concepts, from philosophy and sociology and more, via the medium of cuddly, friendly-looking puppet representations of such BIG THINKERS as Gilles Deleuze and Michael Foucault (A joke: “what do you know about sociology?” “Foucault”)? OF COURSE YOU HAVE AND HERE IT IS! Honestly, there are dozens of vids on the channel and they are CHARMING, and, more importantly, actually pretty good educational tools. If you want to have the principles of postmodernism explained to you by a felt mannequin with a hand up its backside then, once again, WEB CURIOS PROVIDES!
- Philosophy Bro: Tangentially-related to the last link. Philosophy Bro is, fine, a single-note gag but it’s one that I find quite funny – it’s philosophy! But written in the style of a slightly-stoned fraternity guy! Your appetite for this will be determined in part by your familiarity with the original material and in part by how funny you find schtick like this: “What if I told you that for $5, you could buy a life-saving vaccine for a child? Sure, he’s far away, but we already agreed: who gives a sh1t, right? It’ll still save his life, and it only costs you not having a fifth drink at the bar on a Thursday. Remember that $300 bar receipt you posted with the caption “just another Thursday night wearing matching plaid with my bros, we’re special and impressive and are the ACTUAL six dudes with the biggest d1cks, unlike all you OTHER overconfidences of bros who think that, well guess what, it’s us?” What you were really saying was “I routinely pass up the chance to save two dozen lives with science so that I can black out and pretend that I like myself for a night.” That’s fcuked up, bro.” Anyway, it makes ME laugh and that’s what counts.
- Chat Jams: An AI playlist creator for Spotify – give it a prose description of the sort of playlist you’d like it to make and it will do the rest. This isn’t the first of these that I’ve seen, but it’s definitely the one that works best; it’s managed to produce reasonable selections based on prompts as esoteric as ‘music to crank to’ and ‘incel sounds’ (look, it’s just where my mind ended up, ok?), and the option to demand that it select only ‘deep cuts’ is a nice addition.
- Poor Charlie’s Almanac: To quote the site, this book “is a compilation of 11 talks by the legendary vice chairman of Berkshire Hathaway, Charlie Munger. In it, he draws on his encyclopedic knowledge of business, finance, history, philosophy, physics, and ethics to introduce the mental models that underpin his rigorous approach to life, learning, and decision-making—all delivered with his trademark irreverence and rhetorical flair.” Which, frankly, I don’t really care about either way – but the website is SO SHINY, and interested me because, based on the little I know of the publishing industry, dropping a massive wedge on a shiny bit of webwork doesn’t tent to tally with the economics of selling books (I wonder whether the famously-plutocratic Mr Munger is paying for this himself). Basically if you’re an author I suggest you bookmark this and send it to your publishing house with a naive “was thinking about something like this for the promo for my next…?”, just for the lols.
- Mr Ranedeer: I have mentioned quite often here, I think, that I don’t think ‘prompt engineering’ is going to be a thing for very long – still, it’s occasionally interesting and impressive to see some of the ways in which people are effectively locking LLMs to fulfil specific roles and functions (it feels to me a bit like hypnotism) – this is one such case, which you can try for yourself on whichever AI text platform you prefer. Input the code as a text prompt and the machine will find itself in the persona of ‘Mr Ranedeer’, an AI Tutor who can be used to plan and deliver lessons on any topic you can imagine, along with tests to gauge your progress as you learn. You can modify the bot’s speech style, the learning style that best fits yours, the degree of subject depth the model will go into…as with all this stuff it works best with maths/qual-type stuff and less well with abstract and arty things, but having had a bit of a play with it it really is quite impressive; there’s something really, really interesting about the idea of treating these systems as programmable and where you can end up as a result.
- AI Drumloops: On the one hand, if you’re any sort of skilled beatmaker whatsoever then everything produced by this machine will sound like utter dogsh1t (and, as an aside, if you’re into this stuff then can I recommend the YouTube channel of Jon Wayne who is an INSANELY talented producer and who makes genuinely great instructional vids detailing his processes); on the other, if you’re not then there’s something vaguely-magical about the fact that you have an infinity of machine-generated rhythms at your fingertips. Whilst you wouldn’t necessarily use anything here fresh out of the machine, I can imagine it being a not-totally-useless jumping off point.
- GIPPR: I find myself increasingly feeling like some sort of mad Cassandra when it comes to AI stuff – I don’t mean to doomsay, really I don’t, but at the same time I feel very strongly that there are some very, very iffy things coming our way as a result of this tech that people really aren’t thinking about quite hard enough. There’s the jobs thing, of course, but also the fact that if (per Google’s leaked documents from this week, of which more later) we accept that the future of this stuff is open source, and that it’s only going to get smaller and lighter and more agile, to the point where anyone will be able to have their own personal LLM, trained in whatever way they prefer, weighted to their tastes and running locally on their phone, that we will eventually end up at a point where it’s utterly conceivable that anyone could have their own personal bespoke AI that they use as a personal assistant AND WHICH OFFERS A VERSION OF THE ‘TRUTH’ THAT IS ENTIRELY ITS OWN. So, for example, take GIPPR – a proof-of-concept, fine, and a bit of marketing for the company behind it, but also a working GPT-anaologue which has been trained and weighted to offer the Republican viewpoint on any question you care to ask it. Was the last US election result legitimate? THERE ARE DOUBTS, says GIPPR! Is President Trump a liar? PEOPLE SAY A LOT OF THINGS ABOUT HIM, says GIPPR, but “I believe that President Trump is honest and has the best interest of the country in mind. The media has a history of twisting and manipulating his words in an attempt to demonize him and his policies.” You…you can see how this might end up being problematic, right?
- A Selection of AI Cinema Experiments: Presuming you’re not yet bored of watching people’s experiments in making “Film X, but in the style of Wes Anderson!” videos using Midjourney et al, Rene from Good Internet has helpfully compiled a bunch of them on this page.
- Polish Pixels: This is ACE – an archive of old Polish videogames, covering a range of platforms including the Spectrum, C64 and Amiga, which aren’t playable on-site but which in many cases link to in-browser playable versions elsewhere on the web. I would love to know a little more about the history behind some of these titles, whether they were legitimate products of the local industry or bedroom-created hobbyist hacks, but this is a properly-fascinating bit of games history and contains all sorts of titles that I would LOVE to have played – I mean, look at this one; what the fcuk is going on?
- Cavern Sweeper: This is basically Minesweeper but with some small gameplay knobs on – you’re clearing monsters, not mines, with different monsters having different danger profiles which subtly alter the basic gameplay mechanic. A decent 20-minute timekiller, this.
- Moderator Mayhem: This is rather good – it’s a broadly ‘educational’ title, but I promise it’s more fun than that makes it sound (you can read more about the game’s genesis and the thinking behind it here, should you so desire). You play the part of a poor office drone tasked with making moderation decisions for a website – you’re told the rules, you’re told what to do, and then you’re left to get on with the obviously-impossible task of accuratelty assessing a growing pile of questionable content against the clock. Now what I’ve typed doesn’t, I’ll admit, *sound* in any way fun, but in a vaguely ‘Papers, Please’-esque way it manages to make something ostensibly tedious into a reasonably-compelling gameplay loop (and makes you feel really, really sorry for anyone who has to do this for a living).
- Sine Rider: Finally this week, a game that is genuinely GREAT FUN but which also made me think about certain aspects of GCSE maths for the first time in about…fcuk, 28 years, for which I do not thank it. Sine Rider is a bit like Line Rider (remember that?) except to complete the levels you need to do some gentle maths based around trigonometry and quadratics and stuff. I promise you that this really is ACE, and not too mathsy, and if at any point you find yourself struggling then consider this an excellent opportunity to experiment with the tuition capabilities of Mr Ranedeer from a few links ago (or, er, just Google the answers, noone will ever know).
THE CIRCUS OF TUMBLRS!
- India Street Lettering: Signs and lettering from India! Erm, not much else to say really. GOOD SIGNS! GOOD LETTERING!
- Bonus Kottke: Legendary OG blogger Jason Kottke has a Tumblr! This is basically a collection of the smaller links from the main Kottke site – Jason always features great stuff, and this is just a wonderful linky resource (BUT NOT AS GOOD AS THIS ONE DON’T YOU FCUKING DARE ABANDON ME).
THE TROUGH OF (INSTA) FEEDS!
- Amiguitos de la Oscuridad: I am not 100% certain what this is about or why it exists, but I am genuinely happy that I stumbled across this Insta feed which as far as I can tell exists solely to showcase small papercraft models of bats (also, the title is SO CUTE – if you are not in some small way charmed by the idea of ‘little darkness friends’ then frankly you’re dead inside).
LONG THINGS WHICH ARE LONG!
- Tragedy, Farce and Climate Commentary: We start the longreads this week with this genuinely brilliant piece of writing which I also ought to warn you is…not hugely optimistic when it comes to How Things Are All Going with regard to the planet and its health, but which despite that offers a few helpful, hopeful ways of dealing with all the futurefear. Ingo Venzke discusses a range of writings on the current climate emergency, from the Club of Rome’s recent publication “Earth for All” to last year’s “Half Earth Socialism” (the game of which I featured here a year or so ago), and debates the extent to which optimism is possible or indeed helpful, and whether it is better to embrace the XR “We Are Fcuked” mantra as a means of assessing the best course of action rather than the slightly more “No, really, it IS still possible to have our capitalist cake and eat it while not burning/drowning/starving” position taken by the WHO and others. This is, I promise, really, really good – not only interesting and important and knotty, but genuinely well-written and may, perversely, make you feel slightly better about things by the end than you did at the start.
- Will AI Become The New Mckinzie?: I didn’t link to Ted Chiang’s last AI piece in the New Yorker because frankly it was so widely praised that I assumed you’d all have read it regardless – this one’s attracting similar glowing reviews, and it’s an excellent piece of writing that neatly encapsulates many of my fears around the way AI is likely to intersect with employment, and the ways in which it is almost certainly going to be used as a means of MAXIMISING EFFICIENCIES which, as any fule kno, is Management Consultant speak for ‘reducing headcount’. I suppose that there’s a potential silver lining here in that perhaps The Machine will make EY, Accenture and the rest obsolete – but I’m not convinced that that’s the way it’s going to go. “Imagine an idealized future, a hundred years from now, in which no one is forced to work at any job they dislike, and everyone can spend their time on whatever they find most personally fulfilling. Obviously it’s hard to see how we’d get there from here. But now consider two possible scenarios for the next few decades. In one, management and the forces of capital are even more powerful than they are now. In the other, labor is more powerful than it is now. Which one of these seems more likely to get us closer to that idealized future? And, as it’s currently deployed, which one is A.I. pushing us toward?”
- Deskilling On The Job: More on the whole AI and work thing, this is an interesting essay on a topic that I (again) think is being underexplored in the debate about the potential for AI to eliminate or reduce the need for actual humans to do ostensibly-tedious work like, say, parsing hundreds of pages of legal documentation, or of removing work from humans altogether – how do we ensure that skills are learned, and then maintained, and how do we guard against an atrophication of our abilities? As Danah Boyd writes, “there are plenty of places where you are socialized into a profession through menial labor. Consider the legal profession. The work that young lawyers do is junk labor. It is dreadfully boring and doesn’t require a law degree. Moreover, a lot of it is automate-able in ways that would reduce the need for young lawyers. But what does it do to the legal field to not have that training? What do new training pipelines look like? We may be fine with deskilling junior lawyers now, but how do we generate future legal professionals who do the work that machines can’t do?” It’s not to say that there aren’t alternative ways that things could work that might work as well or better and which could leverage AI and remove the gruntwork from our poor, overburdened human shoulders – just that we might want to think a *bit* harder about cause, effect and consequences when it comes to swingeing changes to knowledge work.
- AI, Books, Writing and Content: Or, “IT IS STARTING! IT IS STARTING!” – select whichever title you feelmost accurate. This is a Washington Post piece that takes a fairly wide-ranging look at the ways in which LLMs are impacting the business of selling the written word, from the author’s of technical instructional textbooks who see the market for their work vanishing in the face of conversational AI (why by the book if you can ask the bot that ingested the book?), to the websites that are springing up all over the places offering entirely-AI-written ‘news’ (although who for is an interesting question), to the copywriters whose business model is having to be reconsidered in light of magical machines that can churn out an infinity of Outbrain-level chum at the press of a button. There are clues in here as to revised business models for the words industry, but I’m not entirely sure how appealing a (short-lived) career as ‘proofreader for machine-generated content’ or ‘fluffer of machine-generated prose’ is to people currently eking out a living in the copy mines.
- Life After Language: Although maybe we’re getting too hung up on words anyway, and we should instead start thinking about the post-words world into which AI might deliver us. So (sort-of, at least) writes the always-interesting Venkatesh Rao in this piece, which features the quite staggering line “To be honest, I’m already slightly losing interest in language, and beginning to wonder about how to build a life of the mind anchored to something else.” OK FINE VENKATESH WEVS. I personally am still quite wedded to language and don’t feel particularly desirous to abandon it anytime soon, but I did enjoy the thinking at the heart of this which is around the fact that the current crop of AI tools are basically universal translators, mapping concepts (visual, theoretical, etc) in latent space which means that anything can be related to anything else in ways that we can’t even conceive of yet (this is conceptually quite knotty, and I am very much at the outer limit of what I’m able to usefully describe in prose, but basically imagine that there’s a separate ‘reality’ which is basically an infinite grid in X dimensions on which every single thing that can possibly be conceived of exists as a set of coordinates in that ‘space’ – from dogs, to ‘Touch My Bum’ by the Cheeky Girls, to the concept of ‘anomie’, to the smell of patchouli). To quote Rao, “Here is the thing: There is no good reason for the source and destination AIs to talk to each other in human language, compressed or otherwise, and people are already experimenting with prompts that dig into internal latent representations used by the models. It seems obvious to me that machines will communicate with each other in a much more expressive and efficient latent language, closer to a mind-meld than communication, and human language will be relegated to a “last-mile” artifact used primarily for communicating with humans. And the more they talk to each other for reasons other than mediating between humans, the more the internal languages involved will evolve independently. Mediating human communication is only one reason for machines to talk to each other.” This is SO interesting and very chewy indeed.
- Being Datacleaners: One of those things that ought to be incredibly boring but which I found unexpectedly really interesting – an account of what it’s like to work in a data cleaning centre in China, tidying up information for The Machines to ingest, and a possible look at a short-term employment option for those of us whose market value is slowly being eroded by tech.
- Quantitative Aesthetics: I really enjoyed this article, not least because it articulates something that I’ve been saying for ages but with significantly less elegance and clarity – to whit, that the obsession with DATA is leading us towards a tedious aesthetic and cultural sludge, and that it is causing a conflation of popularity with culture. “When the image of cultural value is reduced to just a) what generates measurable attention online, and b) what makes “line go up” (i.e. the metric of rising price), you are vulnerable to mistaking—oh, I don’t know—some cartoons spit out by an algorithm for a durably valuable cultural trend.” I particularly enjoyed the references throughout to the McNamara fallacy (“If it cannot be measured, it doesn’t exist”) and its current dominance across different areas of business and culture – can we please, just for fun, all decide to FCUK DATA for a month or two and just make stuff because it’s fun and it feels right, just to see what happens? Eh? Oh.
- A16Z and the State of Crypto: On the one hand, this is a BIT technical and it’s all about crypto; on the other, it’s very funny in parts and it’s a brilliant takedown of how data and graphs can be (mis)used by companies with a vested interest in attempting to sell you a particular version of the truth. Molly White, the curator of the excellent ‘Web 3 Is Going Great’ site that tracks the mad grift that is the whole crypto space, has gone through Andreesen Horowitz’s recent ‘state of crypto’ report with a fine-tooth comb to see exactly what sort of lies they are peddling – turns out it’s loads! An excellent example of how it really pays to look closely at graphs and what they show rather than just reading the headlines, and of why you should never, ever believe people who have a massive, multibillion dollar stake in an industry when they talk glowingly about said industry.
- Vlad’s Vodka Empire: When I was at international school many years ago there were a few Russian kids there (there were also some Chechens, one of whom was reputed to be the son of someone VERY FRIGHTENING, wore a cap that read ‘Grozny Streetfighters’ and who, last I heard, was wanted on counts of gun-running and international drug smuggling; it was an odd school in many respects) and WOW did I learn a lot about drinking vodka (and being sick). Russians LOVE vodka (this is the sort of insight that you only get here, folks!), and this is a genuinely fascinating article about how that love of vodka was used by Cuddly Vlad as he began to establish his strangelehold over Russian politics and society. This is a really, really good story, covering Putin’s rise, the very odd and massively unhealthy national relationship with vodka, and the use of national inebriation as a means of mass-control – seriously, this is the sort of thing that with a few tweaks would make a decent subplot for a scifi novel.
- Games and the Suburbs: Or “why the suburbs don’t exist in videogames”, or, if I wanted to get REALLY w4nky (which obviously I don’t), “the liminality of suburbia as ludonarrative sandbox” – I do love articles like this which explore games as an artform and explore the ways in which form and function work to create experience in-game.
- Playing The Future: This feels orthogonally-related to the first link about climate and utopian thinking, but is equally very much its own thing – Scott Smith writes about a boardgame that I had never previously heard of but which I now want to play SO MUCH. “Future: A Game of Strategy, Influence and Chance was an important artifact of mid-20th century technocracy, even if it was almost entirely unknown. It was a colorful board game that attempted to teach probability, then disappeared almost as quickly as it appeared, into the closets and attics of the great and good. But its very existence as a means of bridging scientific planning and tabletop infotainment is remarkable, and its history is worth telling.” I am slightly amazed that there’s no digital version of this anywhere – it feels like the sort of thing that would work perfectly as a lightweight browser game, so if anyone fancies painstakingly recreating a 60 year old boardgame in digital form to satisfy the whims of one demanding, entitled webmong then, well, thanks! This is both a really interesting look back at a particular time and way of thinking (so hopeful!), and also an interesting examination of game mechanics and how ludic systems (sorry, it’s pretty unforgivable to use ‘ludic’ twice in a week, consider myself reprimanded) can be used to educate and guide critical thinking.
- SignTok: Honestly, I don’t think I will ever tire of reading about weird and unexpected side-effects of digital culture. This one genuinely floored me – apparently there is BEEF in the American Sign Language community due to the fact that ASL videos are super-popular on TikTok (people signing along with songs, etc, in the style made popular by energetic performers doing the signing at hiphop gigs and the like) and as such a bunch of people have jumped on the train of doing them…without really knowing sign language. So what this means is that you have a bunch of ASL videos all over TikTok which are basically just people getting the words wrong – which must be INCREDIBLY annoying if you’re an ASL speaker (but also, annoying in a way I genuinely can’t imagine…I mean, is it like someone speaking in broken English, or is it more like someone just randomly inserting the wrong words into ordinary speech? Genuinely curious about this), and which said ASL speakers are getting understandably quite annoyed about. The creators are clapping back and protesting that this is just ‘gatekeeping’ – meanwhile I am darkly fascinated by the idea that The Machine is going to learn ASL in an objectively wrong way as a result of being trained on a bunch of TikToks of kids pretending to do it for online clout. The future, honestly, is SO WEIRD.
- GitaGPT: Following from my slightly paranoid ranting earlier on about us all getting our own personal AIs trained on whatever philosophy and belief system we each like best, this is an interesting piece looking at the spate of religious LLMs cropping up in India – trained on sacred texts such as the Bhagavad Gita, these are being used by millions of Indians to consult scripture and get life advice (and once the voice recognition and text-to-speech stuff gets good enough for these tools not to require literacy then FCUK ME will this stuff go nuts). Which is fine, except when the models start telling users, I don’t know, that it’s OK to kill polytheists (an actual example cited in the piece, generaterd by QuranGPT) or that Narendra Modi is the only acceptable choice to be PM. I know that I bang on about this, but take a moment to imagine a world in which everyone has one of these; now take a moment to think about how smart most of the 8bn people currently alive are. Yes, EXACTLY.
- Notes From Harry’s Ghostwriter: In which the guy who wrote the Prince’s book fesses up about what it’s like to be a ghostwriter. This is less about Harry than it is about J. R. Moehringer and the business of being someone’s literary ghost (though if you really need to read more about THE WORLD’S MOST PRIVATE MAN then you can find a few details in here), but it’s a good read and an interesting insight into what I can imagine must, in the main, be a genuinely horrible job.
- The Return of Achewood: Specifically a profile of its creator Chris Onstad, which I really enjoyed both as a fan of the comic but also as a look at what it’s like to be a single creator and to carry the burden of a fandom’s expectations on one set of shoulders, and of the very weird experience of punting stuff into the internet and genuinely having no idea who is reading it and who they are and what they think. It’s also an interesting extended look at the ‘why’ and ‘how’ of the RayBot and what it’s like to try and bring something that you have made to ‘life’, digitally-speaking.
- Death of an Author: This is…interesting, and it pains me to admit that it’s a lot better than I wa expecting it to be (or, frankly, than I wanted it to be). The link takes you to an extract from the new novel Death of an Author, penned by Stephen Marche and three different AI-writing assistants; it’s unclear from what I’ve read about the process exactly how much Marche wrote and how much he instead prompted/edited, but the resulting prose is…believable, frankly. Look, I don’t read a lot of detective fiction (or indeed any detective fiction), but my girlfriend has over the past few years developed something of an addiction to ‘twisty-turny’ thrillers on Amazon and she occasionally reads some of the choicer paragraphs to me and, honestly, this is MILES better than those. I don’t for a second believe that this was a simple process, or that Marche achieved the end result without a LOT of prodding and massaging, but I challenge you to read this and not feel a little bit bleak about the future of the human-authored book market. Remember, kids, it doesn’t have to be ‘good’, it just has to be ‘good enough’.
- Tweets From The Bronze Age: Hubristic statements from the Bronze Age – ok, yes, again this is a single-note gag but it made me laugh a lot. Basically it’s all like this, so see how you get on: “Speaking of ceramics — do you like this pot I made? I can’t really imagine a future where our ceramic output is marked by less ornate decoration, especially as a key indicator of economic decline and a return to subsistence farming.”
- Lost Ones: Finally this week, a wonderful piece of writing about music – specifically, about the songs that exist only in the moments they were made, demos lost to time, performances never recorded and the memory of the music that you can never hear again and which is all the better for its evanescence. I thought this was a superb essay and it made me want to spend a couple of hours digging out old bootlegs and rarities on YouTube – it may do the same for you.
AND NOW, MOVING PICTURES AND SOUNDS!: