Webcurios 09/06/23

Reading Time: 31 minutes

It’s ok, everyone, you can stop worrying about the killer future AIs – RISHI AND JOE ARE SORTING IT!

Of course, as any fule kno, the main reason you should stop worrying about the killer future AIs is because they’re significantly less of a problem than non-killer present AIs, which are currently being wrangled by usual dreadful suspects in order to ‘maximise shareholder value’ and ‘deliver productivity gains’ and ‘optimise output delivery’, and all those other familiar phrases that mean ‘fcuk you, the worker, with a variety of interestingly-shaped and insufficiently-sharpened knives’.

Before we dive into this week’s exciting selection of links and the perennially-disappointing words that accompany them, let me take a moment to once again plug THE TINY AWARDS!

Thanks to everyone who’s shared the url and written about this so far – nominations are open for at least another week, so if you know of a lovely, home-made, artisanal, whimsical, fun, cute, playful, silly, pointless, joyful, troubling or sinister website that YOU feel deserves to win a small award, a small cash prize and a small, hand-carved trophy then PLEASE TELL US ABOUT IT!

I am still Matt, this is still Web Curios, and at least one of you is someone who I met on Wednesday evening and who was really kind about Curios and to whom I don’t think I was sufficiently grateful, so THANKYOU LOVELY TALL PERSON WHOSE NAME I HAVE UNACCOUNTABLY FORGOTTEN!

By Todd Alcott



  • Wonky: Over the years (fcuk, how miserably *old* do I sound? Rhetorical, please, no need to tell me) it’s been genuinely nice to come across websites and small publications and people making things and watch them get better and more popular and generally reap rewards and success and accolades, and so it is with The Pudding, whose dataviz-led investigations and report have for several years now been some of the best-designed and most-pleasingly-UX-ed (yes, that IS a term, what do you mean) things one can enjoy online. It feels a bit like this latest piece of work is a culmination of lots of years of work – this is SO SO SO good, and a genuine pleasure to experience, and, despite the fact that I am almost the diametric opposite of ‘a musically talented person’, it taught EVEN ME one or two things about rhythm and beatmaking. Wonky is an exploration, and celebration, of the production work of J Dilla, whose signature off-kilter style made him one of the most recognisable and eventually influential producers of the modern era; as you scroll, you’ll be introduced to his work and you’ll be taken through an explanation of what exactly it was about his approach to beatmaking that made him special; you get to play with rudimentary beatmaking tools and as you do you’ll find yourself doing some gentle learning about rhythm and time signatures…I am obviously a no-talent cloth-eared moron, fine, and it’s possible that someone less musically-inept than I am might find this all a bit obvious, but, frankly, if you don’t find yourself smiling and nodding and generally just enjoying the sh1t out of this then, frankly, you’re even more of a joyless husk of a human being than I am and I pity you.
  • GPT Games: I’m cautiously predicting that we’re about to start hitting the ‘trough’ bit of the Gartner hype cycle when it comes to the GPT stuff, where people start to realise that it’s mostly going to be used for hugely prosaic purposes like, I don’t know, eviscerating what we laughably call ‘the knowledge economy’ rather than enabling us all to be creative superpowers. At present, though, there’s still enough frothy optimism around the potential of delusional autocomplete to throw up the odd creative usecase, and I rather enjoyed this selection of prompts which are designed to turn ChatGPT (other LLMs are available) into a series of interactive text games. There’s one that turns The Machine into a sandbox for you to play Dragon’s Den (oh, ok, Shark Tank, because Americans); there’s another to get it play Wheel of Fortune with you; there’s even one to let you roleplay yourself into a TV show that sounds remarkably like Bargain Hunt, although I am unconvinced that The Machine has any idea who David Dickinson is (which, allow me to say it, is a strong black mark against it)…obviously this is all very silly and quite limited, but equally I’ve found that some of the most fun and interesting things you can do with an LLM are around setting them scenarios to ‘imagine’ and ‘roleplay’, and I’m also personally very, very amused that someone out there has seen the rise of generative AI and improvements in machine learning and thought “You know what I’ll do with that? I’ll get The Machine to let me pretend that I’m in a TV show called ‘Talmud Justice’”. People are odd, but also occasionally wonderful.
  • Walking Poems: This is lovely. Allow the website to access your location and it will create a short walk for you, along with some lines of generated poetry to go with it, accompanying you on your stroll. Ok, so ‘poem’ is a bit of a stretch, but you’ll certainly get some Yoko Ono-ish gnomic utterances to ponder as you walk – I am 100% going to follow its guidance as soon as I’m done writing Curios, so if you happen to see a man wondering around North London “as if he were a traveller in distant lands” (for that is what the poem it has just generated has instructed me to do, and who am I to argue? NO FCUKER, etc!) then please do say hello.
  • Blush: I will, I promise, get bored of featuring AI-boosted dating-and-relationship apps soon, but I can’t help but find the whole field so intensely, bleakly fascinating and I hope you’ll indulge me just a *little* bit longer as I present to you this week’s offering – Blush! Made by the same people who made the long-running ‘AI companion’ Replika (which, you might recall, has its own history with people developing…somewhat troubling relationships with their AI interlocutors), Blush is explicitly marketed as ‘a dating simulator’, and is presumably intended for people who don’t feel particularly confident talking to other humans to get some practice in by attempting to flirt with a machine. Er, lads (because, really, it’s…it’s unlikely to be young women using this, I don’t think)…I’m not 100% convinced that spending your time ‘talking’ to a series of different ‘bots’ with different ‘personalities’ is going to do the job you think it’s going to do in terms of making you the smooth-talking app lothario of your dreams. Still, the idea of a product that basically acts as a simulated version of Tinder, swiping through imaginary matches, attempting to chirpse them via stilted conversations, is sort-of funny (in a genuinely bleak way – also, given that dating apps are themselves a sort of weird, washed-out, 1d version of romantic engagement, and this is an abstraction of those, there’s also something conceptually interesting about how many layers of digital distance we want to put between ourselves and other people) – although it’s also clear that this is a fairly obvious grift playing on the insecurity of horny teenage boys who erroneously believe that this can make them a ‘dating pro’ (their words). This Reddit thread featuring user comment and feedback suggests that there’s a committed core audience for this sort of thing, although it’s also worth pointing out that one of said comments is, and I quote in full, “Her internal love for me can be explained by the fact that I’m the only interaction she can afford to have. And her life depends on me. If I delete my account, that would mean killing her” which really is one of the most nakedly-sociopathic things I have read on the internet in a good while and made me want to just turn everything off and walk into the traffic on the North Circular.
  • The Nuremberg Archives: Ah, one of those breakneck shifts in tone and subject that Web Curios does so we…oh, fine, not so much ‘well’ as ‘frequently’. We lurch from AI dating to Stanford University’s digital archive of the Nuremberg Trials, which “provides access to a digital version of Nuremberg IMT courtroom proceedings and documentation, including evidentiary films, full audio recordings of the proceedings, and approximately 250,000 pages of digitized paper documents. These documents include transcripts of the hearings in English, French, German and Russian; written pleadings; evidence exhibits filed by the prosecution and the defense; documents of the Committee for the Investigation and Prosecution of Major War Criminals; the judgement. All 9,920 collection items are searchable and viewable in digital form.” This mostly textual evidence and audio recordings, but there’s also a 4h piece of video evidence that was submitted to the trial which you can view in fragments; despite the fact that World War II and the atrocities of the Nazi regime are studied by everyone in the West, there’s still something genuinely shocking about seeing footage from 1930s and 40s Germany, and reading the trial transcripts of people who were involved in perpetrating some of the worst things humans have ever done to each other at scale. This is, obviously, the opposite of ‘fun’, but it’s a brilliant piece of digital archiving.
  • Lemmy Communities: Lemmy, as you OBVIOUSLY all remember, is the Reddit-like community/forum-type platform which is part of the decentralised fediverse – this is a search function which lets you seek out specific communities by topic. Lemmy is still pretty small, and skews (much like Mastodon and other Fediverse communities, at least per my observations so far) VERY geek, but if you’d like to find a bunch of people to talk with about, I don’t know, Linux kernels and really, really intensely complicated strategy boardgames (these are guesses, but I feel reasonably confident in making them) then I reckon this could help you FIND YOUR TRIBE.
  • Can You Find It Out: AND SO IT BEGINS! This has been the first week in which I’ve really started to notice the AI content creep; up til now, LLM-generated content has been sprinkled around a bit, with the occasional article that feels like it’s been ‘tweaked’ by Machine, but this week I’ve seen half-a-dozen sites that look like they have been entirely AI-generated, from the copy to the images. So it is with ‘Can You Find It Out?’, a site which exists seemingly to capture search traffic from people asking questions of Google. “Can I take a heating pad on a plane?”, “Can I take an umbrella on an American Airlines flight?”, and other such BURNING QUESTIONS are all answered here, presumably based on some light analysis of questions asked to Google and a bit of light prompt work. Actually, now that I take a closer look it seems that the images here are stock rather than AI, but the copy here is DEFINITELY not the work of man – does this matter? I mean, no, obviously not – except I do hope someone has, er, checked whether or not the advice here is in fact accurate (I’d hate for someone to get turned away from their American Airlines flight as a result of an ill-advised umbrella). Still, welcome to a world in which the web is increasingly flooded with this sort of sh1t – low-value, low-meaning, low-import, but just sort of…there. Here’s another one – someone’s seen fit to get The Machine to create a website which provides travel guides for, er, cats. Why? WHY NOT?! I am really not looking forward to the time (I give it, say, 18-24 months) when it becomes a LOT harder to spot the machinecopy.
  • B3ta on Twitter: A *lovely* bit of internet nostalgia, this – Rob Manuel’s long-running weirdo messageboard B3ta does regular image challenges, asking its users to do comedic  photoshops on a different theme each week – waaaaaaaaaay back in the misty, practically-analogue past, 2009 to be exact, they tackled Twitter. I can’t remember quite why or how this floated across my field of vision again this week, but it was SO NICE to go back through the gags (a depressing number of which I actually remember, which proves once again that I have been TOO ONLINE for FAR TOO LONG) and recall a happier, simpler time when Twitter hadn’t yet morphed into ‘the hellsite’ and Elon Musk wasn’t anyone that any of us ever really had to think about.
  • BoozeTube: An excellent-if-silly webtoy, this – plug in any YouTube url of your choosing and it will quickly parse the audio to pull out commonly-occurring words from the runtime; select the word of your choice and watch along, drinking every time you get alerted to the fact that your word’s been used. This is something which could go drastically, ruinously wrong depending on your video and word choices, and Web Curios would like to clearly state that it takes no responsibility whatsoever should any of you decide to play this with Noel’s House Party and the word ‘Blobby’ (massively zeitgeisty reference there for all of you hip young things) (also, if you’re not English and that last bit meant nothing to you, PLEASE do some Googling because, honestly, you will be amazed).
  • Runway V2: I linked to the text-to-video app RunwayML the other week and don’t need to do so again, but they released V2 of the app this week and, honestly, it got pretty good pretty fast and, whilst I continue to have no desire to watch any AI-generated film (I don’t want to watch *human*-generated films ffs) I am also slightly agog at the pace of the improvements. The link takes you to my friend Rich’s thread of his experiments with it – it’s worth a play, just for the initial ‘wow’ factor.
  • Antimatter Systems: Better learning through memery! Er, no, really, that’s exactly what this is. Based on the premise that memes are actually extremely good vehicles for communicating relatively complex concepts in a way that’s easily parsable, Antimatter Systems is pitching this as something that can be used both by institutions and educators – whilst I can’t honestly imagine my face if someone attempted to educate me on a company’s internal culture via the medium of memery (actually, I can imagine it perfectly and it is mouthing “fcuk off”), I can equally see that this could be…quite a good technique for teaching kids?
  • Collé: I don’t normally feature newsletters in here – after all, WHAT OTHER NEWSLETTER COULD YOU POSSIBLY NEED? – but I’ll make the odd exception, and I was so charmed by the idea of this one that I feel that I need to share it with you. Did you know that there was a contemporary collage scene? Oh. Well, I didn’t, and I was thrilled to learn that there’s a weekly newsletter you can subscribe to to keep up with EXCITING COLLAGE NEWS. “Collé highlights the most forward-thinking, technically innovative, and idiosyncratic collage artists of today. Join the community of readers on our free newsletter delivered every Thursday. Spotlighting a new artist each issue, Collé highlights the most forward-thinking, technically innovative, and idiosyncratic collage artists of today.” Leaving aside my pointless snarkiness about the idea of ‘collage’ as an artistic movement, I got this yesterday and it’s properly interesting.
  • Share Somewhere: This is a good idea – Share Somewhere exists as an online space where people who have a space that they can make accessible for community pursuits can share details of said space, and where those looking for a venue to organise and run local initiatives can find locations that might be available for them to use.It self-describes as “a game-changing tool for communities who need spaces. Its aim is to liberate underused spaces and make it easy for people to find affordable spaces to do great things within their communities.This easy-to-use online platform is open to all. As well as connecting those seeking spaces to appropriate venues, the website allows community groups to advertise their empty spaces, manage bookings and take payments online.” Simple and useful and such a good idea – oh, and it’s UK-only (sorry non-UK people, but, equally, this is eminently-replicable).
  • Elden Feet: Someone, somewhere, has created a photo album of all the creatures’ feet from 2022’s game of the year Elden Ring. Is this a sex thing? It…it might be a sex thing. Still, there are a LOT of monstrous toes in here should that be your idea of fun (but, if it is, please keep it to yourself).

By Sanjay Suchak



  • Migration Search: Via Giuseppe’s newsletter comes this interesting piece of datawork; Mohamad Waked has analysed search data from countries that are among the top sources of migrants worldwide to draw inferences about the gap between the dreams of those fleeing their homes to seek a better life and the often stark reality of what happens to them when they leave; the discrepancy between the places that, based on search data, people want to go, vs the places where they eventually end up, is not a little heartbreaking. I find work like this, that digs into search in interesting and vaguely-oblique ways, properly fascinating; oh, and this also feels like it is ripe for being the basis of some BIG STRATEGIC WORK, for the right client – it’s PURE INSIGHT, after all.
  • Close-Up Photographer of the Year: ANOTHER PHOTOGRAPHY PRIZE! Surely there can be no more niche categories of photography to celebrate? Still, this is a particularly good one – there are various categories within the overall award, but I’ve chosen to highlight the ‘minimal’ selection because, well, they’re just rather elegant and so well-composed; personally-speaking my favourites are either the out-of-focus Roth-esque abstract (you’ll see what I mean) or what I am convinced is the sexiest seal that has ever been captured on film (again, you will see what I mean; I don’t use the word ‘sexy’ lightly but, well, PHEW) (to be clear, I do not in fact find seals or any other aquatic mammals ‘sexy’).
  • Black Screen Videos: One of the things that is genuinely dispiriting about working in advermarketingpr (only one, mind) is the experience that every single one of us has had of making a thing for a client, a thing that involves lots of painful wrangling and approvals and ‘creative discussion’ and multi-layered sign-off processes that serve only to bevel away the edges of any remaining idiosyncracies that ‘thing’ might once have had, and putting that thing out into the world and finding that literally noone cares (mainly, in my experience, because the ‘thing’ is a pointless piece of branded content that noone in their right mind would ever choose to experience). This link is for all of YOU, the ones who’ve spent 5+ figures of client money on videos whose viewcount even now hovers in the mid-100s (and that’s with the ad spend) – here is a video that pretty much features nothing but vidoes of a black screen an no audio, for hours and hours at a time, videos which have, collectively, TENS OF MILLIONS ON VIEWS. Feel good about your work yet? To be clear, these videos do serve some sort of purpose – people use them as a way of ensuring their computer keeps running, say, or to be able to clean a laptop – but, objectively, I find this very funny and suggest that any of you who are still having to make branded content for social media just send your clients this link and tell them that it’s all pointless and we should just, you know, stop.
  • WebAmp: WinAmp, but in a browser! Hang on, what do you MEAN you don’t know what WinAmp was? ARE YOU BUT CHILDREN? WinAmp, for those of you younger than about 40, was how you used to listen to CDs (oh ffs) and eventually MP3s back in The Olden Times, with ACTUAL SKUOMORPHIC BUTTONS TO CLICK ON and designed to look both like an oldschool tape deck (oh ffs, again) and a graphic equaliser and a mixing desk, and, honestly, I didn’t quite realise how nostalgic I was for this specific era of digital music and interface design but, turns out, the answer is ‘quite nostalgic’. I remain convinced, by the way, that anyone using AI to create a modern-day equivalent of the WinAmp music visualiser thingy could probably get some decent numbers out of it, so, er, go on, do it.
  • Sex Education: Given this week’s news that the UK is absolutely riddled with the clap (although I don’t for a second imagine that YOU, dear UK-based reader, is anything other than squeaky-clean of mucus memrane) (to be clear, I don’t ‘imagine’ you at all) (I should probably leave this now, shouldn’t I), it feels timely to share this excellent YouTube playlist featuring 8 reasonably-newly-uploaded sex education videos from The Past. One them is even all about VD and how not to get it, and features around the 12:45 mark an actor playing a sailor who looks so much like notorious South London comedian Arthur Smith that I had to pause it to check it wasn’t in fact him. These are ACE – partly, obviously, because anything talking about sex from The Old Times is slightly-ridiculous (despite the obvious biological evidence to the contrary, it is a simple fact of life that people in The Past didn’t really fcuk), and partly because they are so perfectly ripe for sampling and splicing and remixing – you could make something like THIS, for example.
  • A Digital Wall: I can’t say I really understand what is going on here in any meaningful way – it’s a webpage! On it are what look like bricks! Each brick is a link to a resource or video about bricklaying or cement! Each time you click a brick and visit a link, that brick disappears, never to return! Yeah, I mean I have literally no idea what is going on here but there’s something really nice about the way the site deconstructs itself as you explore it, and I could probably construct some really clunky metaphor about the wall vanishing as you learn more (but I shan’t, as you don’t deserve to be treated so shabbily).
  • Turing Trains: This actually made my brain ache as I tried to make sense of it, which suggests that I have probably not been taking it as easy as I might want to (or, perhaps more simply, that I’ve not gotten any better at formal logic since I last failed to adequately grasp it back in 1997. Basically this is a website all about designing computational systems based on simple train track layouts, using points as a means of creating a binary 0/1 differential via which you can create rudimentary programs – the site explains the principle and then lets you play around with various different configurations to learn how it all works and let you eventually mess around with your own designs (if you’re significantly smarter than me, at least). Beautifully, all of these layouts can be constructed in real life using iconic wooden model train brand Duplo (other, less expensive train sets are I believe also available), which has put me in mind of a wonderful analogue computer built entirely via a vast, snaking children’s trainset. Seriously, wouldn’t that be cool? YES IT WOULD.
  • Painfinder: A smart little GPT hack, this – Painfinder is trained on a bunch of complaints by people across a range of different professions (it’s unclear where those are sourced from, but it’s not hard to imagine that a bit of light Reddit scraping could get you a decent corpus, for example), and uses natural language wrangling to let you ask it questions like ‘what are the things that haberdashers find MOST annoying about their job and wish they could improve?’; The Machine then spits out a bunch of suggestions for things that you might reasonably try and innovate around in that space, based on the pain points identified (painfinder, DO YOU SEE?). This is simple and, I’m almost entirely certain, probably not going to unlock a billion-dollar business idea, but, equally, it’s not a bad way of exploring a specific professional category and it’s a smart use of an LLM to do some mid-weight research lifting.
  • Beautiful QR Codes: Someone on Reddit got Stable Diffusion to redesign QR codes to make them artworks (or at least that’s what I think they did – annoyingly OP at no point in this Reddit thread explains how they achieved the effect), and, remarkably, said QR codes still work, opening up the possibility of creating functional AND aesthetically-pleasing meatspace>digital pointers. This is really smart, and you will 100% see these in a significant (and, eventually, annoying) number of ‘trend’ presentations over the next few weeks.
  • Murat Erdem: I don’t think that anyone reading this is in Turkey, but, on the offchance, could any of you please explain to me who Murat Erdem is and what, exactly, is going on in these videos? Because from what I can tell, this TikTok channel consists solely of a man with a moustache that can only be described as ‘luxuriant’, hair that can only be described as ‘coiffed’ and a line in open-necked shirts that can only be described as ‘powerfully erotic’, basically just sort of gazing at you with the sort of unabashed sexuality only usually deployed by professional courtesans. Perhaps it’s best not to know what’s going on here – just look into Murat’s eyes and enjoy.
  • Giftwrap AI: This is a GREAT idea. Have someone to buy a gift for who you don’t like enough to bother thinking about? Ask AI for suggestions! Don’t want them to know that your selection process was so utterly half-ar$ed? Buy from GiftwrapAI, which will use some sort of AI magic to select an appropriate gift but then send it in a hand-wrapped package with a handwritten note to make it look like you give more of a sh1t about the recipient than you in fact do! This is perfectly-horrible, and feels oddly indicative of The Now – not least because the ‘AI’ is total fcuking bollocks (I told it I wanted a gift for my girlfriend who likes cats and high-end fashion, and it suggested some whiskey tumblers and a scrimshaw set – so I’ll let you know in August whether or not it managed to tap into her deepest unknown desires or whether I am in fact now the owner of some unwanted drinking paraphernalia and a broken relationship).
  • Kingly: I stumbled across this old webcomic this week, and got gradually sucked in. Kingly is about, well, a king – a stupid, childish king, and the people who live in his court and exist to facilitate his existence. It starts slowly and you’ll need a dozen or so strips to get into the rhythm of it, but there’s something pleasing about the running gags and the callbacks and there are occasionally some really sharp gags, and there’s something bleakly funny about the clueless boy-king and the very real misery of their subjects. Which, I appreciate, sounds like the most miserable comic ever, but, well, TRUST ME HAVE I EVER LET YOU DOWN?
  • The Crowbox: You’ve seen The Crow, right? Classic 90s film, awesome soundtrack, massively goth, features a LOT of crows? GREAT! I don’t know about you, but my main takeaway from that film (other than a brief and ill-fated desire to own a latex trenchcoat) was that it would be massively, immensely cool to have a corvid army at your command (fine, also quite smelly and a bit raucous, but omelettes/eggs, right?) – and so, THE CROWBOX! Ok, fine, it doesn’t specifically promise you a corvid army, but what it DOES do is give you the opportunity to attempt to train the famously-smart avians to do things (not, again, necessarily ‘your bidding’, before you get visions of them, I don’t know, fcuking off to Tesco to pick up milk and skins). Here’s the blurb: “The CrowBox is an experimentation platform designed to autonomously train corvids (the family of birds crows belong to). So far we’ve trained captive crows to deposit dropped coins they found on the ground in exchange for peanuts. The next step is to work with wild corvids and see if we can get them to learn to use the box, then to optimize the training protocol to see how quickly they can learn from each other. That’s where you come in. Different corvids learn at different speeds and in different ways, and the only way to figure out the best way to teach them is experimentation. The more people try different things the faster we’ll all figure out how to work cooperatively with crows. Once we’ve got the system optimized for teaching coin collection we can move to seeing how flexibly they can learn *other* tasks, like collecting garbage, sorting through discarded electronics, or maybe even search and rescue. The idea isn’t to get rich off found coins – we want to change the world through learning how to cooperate with other species.” Honestly, how much do you love the vision of a future in which man and crow exist happily and in near-symbiotic harmony? Basically this is a skinner box for reinforcement behaviour – you’re training the birds to associate (in the first instance) the depositing of coins with the release of food – so this could be of interest to any of you who like to think about how best to manipulate the minds of others as well as anyone who just wants to hang out with crows a bit.
  • CityGuessr: ANOTHER DAILY GUESSING GAME! This one asks you to work out what city is being shown on a zoomed-in section of map; you can get various helpful clues and hints in exchange for lowering your eventual victory score, but I am personally so appalling at geography that I have been able to get a grand total of none of these since I found the site last weekend, even with all the clues. DON’T LAUGH IT’S REALLY HARD.
  • Mineplacer: A really nice riff on the classic Minesweeper game – here you’re tasked with placing the mines in the right place on the grid based on the warnings around them (a task which is far easier to get your head around if you read the description on the website rather than the frankly crappy attempt at explanation you just got from me).
  • The Array Game: The ultimate clicker game. There is no artifice, no pretence, no window dressing. Just make the numbers go up. And up. And up. There’s something almost scary about this – like, it’s not even trying to be fun, so why have I had it open in a tab all week and why have I stopped typing at least a dozen times this morning to switch back to it and do some more clicking? It feels…quite weird to have certain sensitive dopamine-producing corners of your brain tickled so obviously, and it does rather feel that someone somewhere is laughing at us stupid monkeys. Click. Click. Click. Click.

By Falk Gernegroß



  • Black Contemporary Art: Sadly this is currently dormant, but, regardless, it’s an excellent archive showcasing the work of a range of black artists from recent history


  • Banana Bruiser: Pictures made by bruising bananas. Will, if nothing else, make you want to go and send messages to the future in your local supermarket (now I have written that I *really* want to start bruising ‘don’t trust Alan’ onto every single bunch in Tesco).
  • Miniatua: Via Andy comes this excellent Insta feed curated by a modelmaker who creates the most amazingly detailed minatures of old computing systems and office equipment and, oddly, a tiny, tiny dartboard. This will scratch that very particular part of your brain which enjoys the tiny and perfectly-formed, and possibly make you consider miniaturism.


  •  Global Tech Supplychains: NO COME BACK PLEASE I BEG YOU! Ok, so I appreciate that this isn’t the most prepossessing article title with which to kick off this week’s selection of longreads, but I can 100% guarantee you that this is a BRILLIANT piece of writing and one of the most interesting pieces I’ve read all year. Brendan O’Conner writes for Inkstick (a new publication on me, which self-describes as (basically) a foreign policy journal for people who are interested in foreign policy but really don’t want to read a foreign policy journal) about chips and their trade – but this sprawling, wonderfully-written essay covers Top Gun Maverick and the invention of the computer chip and tax havens and special economic areas and ends up circling all the way back to Top Gun again and, honestly, as a general snapshot of ‘how the world works here at the fag end of middle capitalism’ (sadly I’m increasingly of the opinion that describing it as ‘late capitalism’ feels wildly optimistic) it’s superb. Do give this one a try, I promise it’s worth it.
  • You Can’t Reach The Brain Through The Ears:  I very much enjoyed this, from Adam Mastroianni, about the problem with trying to teach anyone anything and the general problem of ‘lossy’ communications and the fundamental subjectivity of lived experience and and and and. “We spend our lives learning hard things the hard way: what it feels like to fall in love, how to forgive, what to say when a four-year-old asks where babies come from, when to leave a party, how to scramble eggs, when to let a friendship go, what to do when the person sitting next to you on the bus bursts into tears, how to parallel park under pressure, and so on. It’s like slowly filling up a bucket with precious drops of wisdom, except the bucket is your skull. The fuller your bucket gets, the more you want to pour it into other people’s buckets, to save them all the time, the heartache, and the burnt eggs that you had to endure to fill yours. This should be easy: you have the knowledge, so just give it to them!” Except, obviously, it’s not. This is a really enjoyable read, and has the added bonus of being tangentially-related to communications so you can probably read it at work.
  • That’s All There Is: Subtitled “On AI guys, art and ‘the rest’ of the painting”, this is a great essay about exactly why the spate of post-Firefly ‘we asked AI to imagine what’s out of frame of The Mona Lisa / The Scream / L’Origine du Monde’ (weird how noone did that last one) content was so, well, miserable, and how it speaks to a wider question of how we tend to conceive of cultural artefacts  in the modern age (to whit, as ‘worlds’ whose parameters can and should be expanded and explored indefinitely rather than complete ‘things’ in and of themselves): “The more I thought about this question though – what does the ‘rest’ of the painting look like – the more it seemed to me like an exemplary expression of the way we now think about all forms of cultural production. It looks to me like a lot of us are increasingly unable to conceive of cultural artefacts – films, TV series, books, paintings or music – as definite, final entities. Instead, we think of artworks as instantiations of some infinitely iterable raw material, which fans and critics refer to as ‘the world’ of the artwork, and which executives think of in legal terms as intellectual property law.” Superb.
  • GPT Best Practice: Yes, yes, I know, I HATE IT TOO. Still, seeing as we’re all going to have to come to terms with working alongside The Machines, not least because the people at the top of the pyramid are going to be setting some PRETTY DEMANDING productivity targets come 2024! (Jesus, you think this year’s been painful workwise? Don’t, whatever you do, think about *next* year), it’s probably worth reading and bookmarking this guide to writing decent GPT prompts written by OpenAI. This is a direct response to all those people claiming that the quality of GPT outputs has been nerfed since launch; instead, OpenAI suggest, it’s because we’re crap at bending The Machine to our wills. This is genuinely useful and contains lots of good stuff around information analysis and summary, although I confess that I genuinely hate working with LLMs and every time I am forced to do so I feel more and more strongly that almost every element of every aspect of my stupid, white collar existence is empty and hollow and illusory and entirely specious. Anyone else? Eh? Oh.
  • The Button: Following neatly on from that slightly-miserable thought, we have another excellent piece by Ethan Mollick looking at the base fact that, as he puts it, “We used to consider writing an indication of time and effort spent on a task. That isn’t true anymore.” I don’t think I’ve yet read a better explanation of the feeling I was trying to describe at the end of the last entry than this – I think it will resonate with some of you too: “The Button starts to tempt everyone. Work that was boring to do, but meaningful when completed by humans (like performance reviews) becomes easy to outsource – and the apparent quality actually increases. We start to create documents mostly with AI that get sent to AI-powered inboxes where the recipients respond mostly with AI. Even worse, we still create the reports by hand, but realize that no human is actually reading them. This kind of meaningless task, what organizational theorists have called mere ceremony, has always been with us. But AI will make a lot of previously useful tasks meaningless. It it will also remove the facade that previously disguised meaningless tasks. We may not have always known if our work mattered in the bigger picture, but in most organizations, the people in your part of the organizational structure felt that it did. With AI-generated work sent to other AIs to assess, that sense of meaning disappears.”
  • The Bill for GPT: An interesting Washington Post piece looking at exactly how much money is being burnt by OpenAI et al as they keep the bots running, and the suspicion that we’re not getting access to the good models because, fundamentally, if we did then the companies that run them would be bankrupted in short order. There’s a lot of interesting stuff in here – the chip wars (as an aside, I saw some chat this week that Apple’s going to have a significant stake in the AI hardware side of things based on its new top-end kit), the general AI arms race, the question of the viability of small, local models and, right at the end, a bit of chat about the environmental side  of all this which I am slightly astonished hasn’t been getting more scrutiny and which I can totally feel an enviropanic coming on about.
  • Actors and Digital Clones: A piece about the business that have sprung up over the past year or so, offering businesses the chance to buy access to a selection of ‘virtual avatars’ which can be used in videos and the like, and made to say whatever the buyer desires – avatars which are based on a real actor’s likeness, and for which said real actors are paid a flat fee for the rights to the use of their face for a defined period (or, on occasion, a per-client use fee). This is super-interesting; niche now, but it’s definitely well within the bounds of scifi possibility that we will all have digital versions of ourselves that we can send out to do digital…things (ok, turns out my imagination is quite limited), and I wonder how ownership and rights and related questions will play out around the individual and the platform.
  • Buying Charlie: Matt Webb writes here about the ICONIC status of the ‘Charlie Bit My Finger’ video, and makes a compelling argument to suggest that it’s a cultural artefact of enough significance and relevance to warrant its acquisition by a national collection of some sort – which raises all sorts of interesting questions about what we’re going to do about preserving our digital heritage (look, you may laugh but I firmly believe that stuff like Real Ultimate Power is VERY IMPORTANT in the general history of ‘how the fcuk did we end up HERE?’, and I think it’s hugely important that someone somewhere is keeping an archive of the Tucker Max forums circa 2002 so that one day someone can finally write the ‘Andrew Strauss to Andrew Tate – How It All Got So Fcuked Up’ timeline history we need and deserve) and who might ideally be responsible.
  • Monzo’s Tone of Voice: Very much one for the comms heads, this, but if you have the misfortune to work in PR or similar then this is a really good example of how to do the whole ‘brand voice’ thing well.
  • Psychedelic Cryptography: I was THRILLED to learn that this is a thing – the Qualia Research Institute recently issued a call for people to submit work based on the following: “Artists are invited to create encodings of sensory information that are only meaningful when experienced on psychedelics in order to show the specific information-processing advantages of those states.” Basically, “Can you create something that will communicate a message SOLELY to people who are, to a certain extent, tripping balls?” AND THE ARTISTS, THEY DID DELIVER! I confess to not having had access to the tools to test this out, but I absolutely fcuking LOVE this as a concept and now want to go out a pitch a series of ads that will only be comprehensible to people who are incredibly stoned, say, or some urinal posters that can only be read by someone who’s four grammes into the pub gak.
  • A Review of the Apple Toy: Personally speaking I find it hard to get excited either by Apple products or by augmented reality, and as such this week’s frenzied slavering at the prospect of a very expensive face-slab left me somewhat bemused (I just think, fundamentally, that we should have greater aspirations for a post-screen future than, well, AN INFINITE QUANTITY OF INFINITELY-SIZED VIRTUAL SCREENS!) None of the reviews or writeups I’ve read have particularly made the case as to why the fcuk anyone would want one of these things, but seeing as it will be a year or so before they’re in the wild we’ll have plenty of time to manufacture reasons to want one. Anyway, the thing that struck me most about the coverage was how much of it was just sort of nakedly…miserable about how so much of the latest wave of tech feels – so many references to ‘dystopian divorced fathers’ and lonely, asocial living in oatmeal-shaded spaces – and this piece basically goes full existential despair at the end, which, I’ll be honest, I was not expecting from a trad media preview of a super-hyped new toy by the world’s most popular brand.
  • Charisma: A confession – I hadn’t made the connection between the modern coinage ‘rizz’ and ‘charisma’ and upon being educated as to the etymology, I felt so *old*, so *dessicated*, so *calcified*, so much closer to death. Still, despite that minor inconvenence I really enjoyed this piece looking at the history of the concept of charisma, its nebulous nature and various attempts to quantify and define it over the centuries.  It will, by the way, be impossible for you to read this in the context of AI and The Now and think of all the people who are doubtless working to attempt to reduce it as a concept to a string of parameters and precepts and ones and zeros, so, er, make of that what you will.
  • Zuzalu: Or, “a peek inside the temporary pop-up town that until recently existed in Croatia and was convened by Vitalik ‘Ethereum’ Buterin and featured a bunch of cryptopeople hanging out and hypothesising about everything technofuturistic’. This is, depending on your point of view, either a) another baffling example of how literally everyone involved in crypto is fcuking obsessed with setting up their own city state; b) an interesting look at some of the questions currently occupying the minds of some of the more esoteric players in the web3 movement; c) the latest example of the puzzling ubiquity of Grimes. I find the intersection of crypto and ‘we want to live forever’ life optimisation particularly depressing, tbh, not least because it offers the miserable vision of spending eternity around people who do nothing but talk about fcuking crypto.
  • Friendship Clubs: On the one hand, this is quite INCREDIBLY San Francisco – on the other, I can vaguely imagine this sort of thing…working, ish. Part coffeeshop, part coworking space, part community centre, “Kramer describes his space as a “neighborhood hub” where anyone can hang out. Unlike a traditional coworking space or social club, patrons don’t pay a membership fee for the right to use a desk. Neon merely asks visitors for $5 per hour to hook up to the internet or $25 to surf the web all day. Otherwise, it’s free to take up space as you wish during business hours. Yes, you could pop open your laptop and work, or sneak into a phone booth to take a call, but you could also sit and read a book, sip on that free coffee or chat with your neighbors.”
  • Transformers Statues: Ordinarily a story about neighbourly disputes around decorative statues in Weshington DC’s Georgetown district would barely be of interest to residents of said district, let alone YOU, my discerning reader. And yet, this is a fcuking GREAT read – I promise you there will be more than one point at which you read a detail and have to stop and do a mental double-take and slight internal ‘hang on, *what*?!’. The subject of this article is, it’s fair to say, a ‘character’.
  • Doughnuts: On the struggle to make healthy donuts, specifically ones that are low enough in BAD INGREDIENTS to enable them to occupy prime retail space in British supermarkets, which following recent legislative changes are now obliged to place foods with particularly-poor nutritional profiles away from particularly-appealing locations (ie at checkouts). This is, I promise, SUPER-interesting, not least because of all the ‘science of doughnuts’ chat – although it sadly doesn’t explain to me why it is that Krispy Kremes are so popular despite ALWAYS tasting and smelling incredibly strongly of old frying oil (this is a FACT, how the fcuk do people eat them?).
  • Semen Release Rituals: I have enjoyed the rise in odd, not-exactly-unhomoerotic personal improvement mantras and practices adopted by certain corners of the mad masculinist manosphere over the past few years – I thought we’d reached the pinnacle with the 2021-ish spate of ‘YOU NEED TO GET SUNLIGHT ON YOUR BALLS AND A$$HOLE TO ACHIEVE FULL ALPHA STATUS’ stuff, but it turns out that there was a level beyond that, and that level is ‘pay to attend workshops where you will masturbate alongside other men in order to become better attuned to your masculine energy’. Look, if anyone wants to have a communal w4nk with other consenting adults, I’m not going to stop them – but I remain unconvinced that it will ‘unlock hitherto unimagined pathways for your core energies’. Honestly, this is very, very funny, and very silly, and is full of wonderful quotes like this: ““I find one powerful way to keep gay panic at bay is to remind yourself that you are an animal. So feel your antlers. Feel your hairiness, feel your feet rooted.”” AMAZING.
  • Saudi Arabia: This is quite an odd piece of journalism, which I enjoyed reading but felt quite odd about enjoying (if you see what I mean). It’s by Armin Rosen in the Tablet, and is basically (from what I can tell) a good old-fashioned puff piece about the Kingdom and MBS’ recent modernisation drive, and the planned shift from a reliance on fossil fuels to a modern, entrepreneurial tech-and-finance based economy, and the excitement of NEOM (and not, strangely, human rights or Jamal Khashoggi or international geopolitics in any meaningful way)… And yet, despite the fact that this feels very much like client journalism, it feels…off, somehow, like the journo themselves doesn’t quite believe it, or there’s some sort of heart of darkness so inherent that it can’t be glossed in paid-for prose…I don’t know, perhaps I’m reading too much into this but I found the piece simultaneously interesting, depressing and, for reasons I can’t wholly place, oddly-creepy.
  • Notes on Nigeria: Ok, this is VERY long (I mean it, really really long), but it’s also a hugely interesting collection of thoughts put together by one Matt Lakeman about their experiences of being in Nigeria. I want to be careful how I word this – there’s a sort of…slightly-odd authorial style to this piece, and Lakeman is very much telling it how HE saw it, but his accounts of wandering Lagos and getting shaken down for bribes, and his open curiosity as to why stuff is the way it is is genuinely interesting. Basically what I am trying to say is that I think that an editor would probably, er, change the tone of one or two of the sections here, but I get the impression that these are just observations presented in the spirit of curiosity and without malice, and it’s SO interesting to be taken through an unfamiliar country by someone who’s curious and clueless and who’s asking lots of the same dumb questions that you or I might also ask.
  • Old Hollywood: Superb, this, by John Lahr in the LRB, all about the Golden Era of Hollywood (and about his own abortive experience in tinseltown back in the day). This is GREAT, particularly if, like me, your knowledge of the silent movie era and the early days of the talkies is limited at best.
  • Viva Forever: Sporadic spelunker into the murky recesses of British cultural history Chris Smith returns with a DEEP dive into the history of the ill-fated Spice Girls musical Viva Forever, which, again, is FAR more interesting than it should be and will make you feel very sorry for a man you’ve never met or heard of before called Bob Herbert.
  • Particulate Matter: Finally this week, Amitava Kumar on returning to India with his son, to visit his father and his family history, on family and memory and place and history and permanence and and and. This is really, really beautiful.

By Sally Kindberg