Friday, 29 November 2024

TURNING THE AIR BLUE

Dick Pountain /Idealog 358/ 07 May 2024 01:32

In my back-room hardware morgue is a black cotton bag, about the size of Santa’s Sack, containing nothing but cables. This unwholesome collection bears witness to a long career in personal computing: RS 232 serial, parallel printer, CAT3 and 5 ethernet, BT telephone, USB, phono, SCART, HDMI, Composite Video, and more. (A long-time fan of dub reggae, seeing it prompts me to put on the Bag-O-Wire album).  

My home/office has been almost wireless for years, ever since the advent of WiFi and Bluetooth, and while that’s generally a very good thing, it’s not quite problem free. I listen to a lot of music and am fussy about sound quality, so I route all my sources through a hi-fi system with vintage Castle speakers. But I don’t only listen to music, I also make it, using several guitars and a small Marshall amplifier. 

My last remaining wired connection was from an outdated LG Smart TV to my hi-fi amplifier, which has Bluetooth but supports only one connection at a time, which was occupied by my laptop for Spotify and YouTube. I tried a Bluetooth dongle in the TV’s headphone output, but the switching involved proved more inconvenient than running a wire under the rug. Recently that TV died and I replaced it with a new one with Bluetooth built in, and the fun began. 

The IT industry has come up with a fair number of dodgy protocols, but by far the most perplexing to me is Bluetooth, which often appears to have a (malignant) mind of its own. In a couple of recent columns I’ve mentioned Albert Einstein’s dislike of Spooky Action At A Distance. Well, to me Bluetooth is Snooty Action At A Distance: it connects only when it feels like it and if you don’t irritate it. (I Googled SAAAD and in addition to the Einstein reference I discovered it also stands for South African Academy of Aesthetic Dentistry, a spooky connection).  

Now having a spare Bluetooth dongle (so cheap they could give them away in cornflake packets) I thought it would be nice to put it in the line output of my guitar amp, so I could practise later at night using my JBL wireless headphones. I paired these but my first test was explosive. The guitar output actually came out of the hi-fi at colossal volume instead, but I was wearing the over-ear headphones which attenuated the sound so much that it took me a devastating while to realise this…

Turns out the dongle is happy to talk to the hi-fi amp but not the headphones. The same happened with the new TV where I could use the headphones, but only by manually disconnecting the hi-fi amp via my laptop (which didn’t work for the guitar amp). I also have a Samsung tablet which I use, among other things, to run a midi sequencer app on which I write bass, drum and occasionally piano tracks to play along to. Shifting the Bluetooth dongle from the guitar amp’s output to an input channel and switching it from Tx to Rx talks to the tablet perfectly! But that same tablet won’t pair with my laptop! I may be getting hysterical! And my teeth may be turning blue through excessive swearing! The problem is twofold. Some devices are picky about what they’ll pair with, possibly due to Bluetooth version clashes, which often aren’t easy to ascertain. But also most small Bluetooth devices have no user interface to make visible what else is going in the Toothosphere. I expect someone will write in to say none of this is a problem if you buy all your kit from Apple, or Sonos or whoever, but I’m not interested in wiring the whole house for muzak, merely to connect the actually-existing kit I’ve selected over the years. 

I’ve reached a reasonable equilibrium. My new Panasonic telly has Dolby sound good enough that I don’t need to Bluetooth it to the hi-fi. The wireless midi sequencer works a treat as accompaniment. And I dug out a pair of wired Sony headphones for late night practise, All this I can live with. 

What I can’t live with is the traumatic experience that buying an up-to-date Smart TV has become. In the few years between my old LG and this Panasonic, a pandemic of streaming services has rendered TV viewing a nightmare, with screen after screen full of horrid thumbnails for trashy programs on crumby networks, almost all of which want you to subscribe. This new set regards watching live TV channels as some kind of perversion which it tries to hide away from you. But that’s a topic for another column, on Cory Doctorow’s concept of ‘enshittification’ and the plague of rent-seeking.

[Dick Pountain is now convinced that Smart TV is an oxymoron] 

Monday, 21 October 2024

CHINA SYNDROME

Dick Pountain /Idealog 357/ 08 Apr 2024 01:09

Unless you live permanently as an avatar in Second Life [does that even still exist?] then it  can’t have escaped you that the world we actually live in has become very much more ‘interesting’, that is dangerous, over the last few years. What with epidemics, wars and deepfakery, the dot-com boom begins to feel as remote as the Middle Ages. A couple of incidents last week pushed me toward these reflections…

The first was the announcement that the US DoJ (Department of Justice) was launching an antitrust lawsuit against Apple Corp, for restraint of trade and monopolising the smartphone market. The DoJ’s complaints all arise from the ‘walled-garden’ attitude that Apple has maintained ever since its very first days under Steve Jobs, with strict control over software development and peripheral manufacture so as to completely exclude smaller companies from entering, and restrain the businesses of their larger competitors. Some of the charges are aimed at contractual restrictions, fees and taxes on the creation and distribution of ‘super apps’ for daily functions like social media, payments, banking and video messaging; hindrance of streaming apps like Spotify, Netflix and Google Photo; restricting crucial API access in the smartphone sector and excessive control over how smartwatches operate on the iOS platform, and control over digital wallets that restricts cross-platform wallets on the iPhone.

I’m very far from being an Apple fanboi: the last one of their machines I owned was an Apple IIe and I was an enthusiastic early adopter of the open architecture IBM PC – which is how I ended up here. Wearing my other hat as a commentator on politics I’m an equally enthusiastic advocate of antitrust regulation, an admirer of the Norwegian-American economist Thorstein Veblen whose theories about conspicuous consumption and the leisure class were influential during the last great bout of US antitrust action in the 1890s which curbed the excessive power of ‘robber baron’ industrialists like the Rockefellers, Mellons, Carnegies, Vanderbilts and J. P. Morgan, ushering in a century of industrial growth and lessening inequality in the US (admittedly helped by two world wars). That progress went into reverse in the late 1970s, to a point where even Forbes Magazine now believes regulatory action is justified and that ”Apple’s business will be fundamentally changed by this lawsuit”, citing the example of Microsoft which “successfully rebounded from its own tumultuous years with regulators and has become the most valuable company in the world, ironically by becoming more open and embracing open source rather than shunning it”. 

If the DoJ were to win this suit they would almost certainly then go after Google, Amazon, Facebook and the rest, but as Forbes also points out the case will run for years – and the Democratic Party has at best a 50:50 chance of retaining the presidency after this November. It’s impossible to predict what a second-term, enraged Trump would do, but pursuing the case may not be high among his priorities. A horrible vision arises of Silicon Valley giants panicked into allying with him to create a doomsday authoritarian plutocracy, administered by AI robot warriors out of a Vaughn Bode comic…  

Which brings me to my second event. I’m something of a fan of the quirky, German, ex-particle-physicist-and-YouTuber Sabine Hossenfelder, who last week posted on her channel a remarkable lament/confession about why she left academic physics. She’s already written a book, ‘Lost In Math’, excoriating her contemporaries for being seduced by pursuit of mathematical beauty away from experimental verification. In the new video she paints a gloomy picture of a profession where young physicists have to tramp the world taking short-term posts, forced to churn out less-than-important papers on topics enforced by credit-stealing superiors, all originality snuffed out. And addiction to ChatGPT and its ilk looks set to make this deadening toil worse still.

And my point? Regardless of whether Biden or Trump is POTUS this time next year US relations with China will continue to deteriorate. America will keep on withholding leading-edge semiconductors from China and attempting to repatriate fabrication abilities from Taiwan to the continental USA. China will continue to build its own semiconductor research and fabrication (even if it doesn’t actually grab Taiwan) and will prevent its best students from winding up in the USA. The decline that Hossenfelder depicts in Western academic physics will therefore become an alarming strategic deficit. I’m not of the party that hungers for war against China, and I’m sceptical whether China and Russia could ally to start such a war, but I recognise that Western physics is likely to depend ever more on Silicon Valley and less on academia, which will make the handling of regulation a matter of some delicacy and diplomacy.  

[Dick Pountain is pretty handy with chopsticks]  

TANGLED COSMOS

Dick Pountain /Idealog 356/ 06 Mar 2024 03:39

For all its faults, YouTube is a resource that offers a quite unprecedented breadth and depth of content – all kinds of music, craft, sport, even streaming movies, and both science and mathematics. OK, much of what it offers is of increasingly dubious, click-baity quality, and that even applies to a lot of the science. I view a lot of YouTube mathematics vids, and while they lack the professional depth you can find on Wikipedia, they often make up for that by exploiting visualisation in original and useful ways. Unsurprisingly the science on YouTube is dominated by cosmology and particle physics, those perennial playgrounds for opinionated nerds who know that Einstein or Bohr were wrong (or right, or whatever). To be sure it’s created a handful of stars like Sabine Hossenfelder and Derek Muller who talk sense and explain stuff well, but the nature of the medium means these get drowned out by the charlatans for most non-scientist viewers.

There are however certain ‘brands’ that guarantee seriousness, and one of the best is the Royal Institution, which gave us Humphry Davy, Faraday and Bragg. I always watch their Christmas lecture, because I like bangs, but this year YouTube offered me as follow-up a lecture by Professor Tim Palmer, who I’ll confess to not knowing before. His lecture, which was about uncertainty and probability, more or less blew my mind, and pushed me to read his book ‘The Primacy Of Doubt’ which merely reinforced that effect. 

Palmer studied mathematical physics and became an expert on chaos theory, before turning that knowledge to practical use in weather forecasting. The huge extension of reliable forecasting in recent years, from a couple of days to around a week, is largely thanks to the method of ‘ensemble’ forecasts he pioneered. Shocked by the infamous failure of the Met Office to predict the 1987 UK hurricane, Palmer realised that the deterministic supercomputer models then in use were prone in rare cases to chaotically wrong prediction, which could be alleviated by feeding the model multiple starting conditions that differ by small additions of randomness – instead of a single prediction this produces a brush-like clump of predictions, the central trend of which is more likely to be right.The theory behind his method involves the Lorenz Equations, which describe the behaviour of chaotic dynamic systems. Solving these equations doesn’t lead to single precise solutions, but to infinitely convoluted fractal graphs called attractors. Palmer’s explanation of Lorenz’s discovery in his RI lecture is superb, and he does it without writing any algebra or calculus on the blackboard, admirably accessible. That however is not what blew my mind: that happened when Palmer turned his method to cosmology. 

In addition to his work in chaos theory Palmer (who is now an Oxford Professor of Climate Physics) obtained his doctorate in General Relativity, so it was no surprise that he has tried to bring the two subjects together. Taking off from Einstein’s theory that gravity is the effect of the curvature of space/time by matter, Palmer postulates that rather than being smoothly curved it may be that space/time itself has fractal complexity, sort of like the Lorenz attractors of chaotic systems. I don’t understand his ‘Invariant Set Postulate’ well enough to attempt further explanation here, but well enough to know that if true it would resolve several of the knottiest problems that currently occupy both cosmology and quantum physics. 

Subatomic particles traversing a fractally convoluted space/time would explain quantum uncertainty without having to invent multiple universes, could offer a theory of quantum gravity more plausible than any of the current candidates, and even do away with ‘non-locality’ and the spooky action at a distance that so annoyed Albert Einstein. 

You’ve probably gathered by now that I’m one of those same opinionated nerds that I was deriding above. I’ve ranted in this column before against people who employ their shaky grasp of quantum uncertainty to declare everything from ‘there’s no such thing as reality’, to multiple universes that provide mechanisms for magic and telepathy. As a humble chemist rather than a mathematical physicist I’m not equipped to refute them in the way Palmer may be in the process of doing. I believe that the very real world that we live in is ruled by thermodynamics, not quantum mechanics, whether that be the crappy battery in your electric car, the catastrophic increase in your utility bills, or those massive perturbations of atmospheric energy that we call our climate (and which may yet see the end of us if we don’t start concentrating). Tim Palmer knows that better than almost anyone, and I have no hesitation in recommending that you watch his RI lecture here https://youtu.be/RkiEV47KPX4?si=BvcXzGzhNqOCiAuv.  [


 Dick Pountain is still just about capable of performing useful work on his surroundings]    

CLOUDY THOUGHTS

Dick Pountain /Idealog 355/ 05 Feb 2024 01:24

Internet culture was once expected to join the whole world together, and in some ways – for example email – it still comes closer than any previous technology. However in other ways it divides us up into radically different camps and silos, especially when it comes to publishing one’s own multimedia content. 

This thought occurred to me recently when I finally succumbed to curiosity by opening a Substack account. I’d been hearing about this service for several years, as used by many people whose work I read, so I decided to put up some of the material from my existing Blogger blogs and website. First impressions were of deep confusion, greater even than that I feel on Instagram. Substack combines a blog for publishing new short material with a website for long-form essays, an email distribution and publication system and a system for getting paid. I find its UI quite opaque right now, and it suddenly flashed on me how many places I now have my own ‘stuff’ online, most of which don’t pay anything at all.I generate content in the following media: text, like this column you’re reading, plus book reviews for other print publications; pictures, photographs and digital art; music, some computer generated, some played, some just curated playlists of other musicians’ work. I currently keep text online on Blogger, Medium, Substack, OpenDemocracy, The Political Quarterly and several smaller publications, and a book published in Amazon’s Kindle store. I have photographs online at Flickr, Facebook, Instagram, Tumblr and my own website. I have music on YouTube, SoundCloud, Bandcamp and my own website plus hundreds of playlists on Spotify and YouTube. That makes at least 20 different places with different addresses and logons, some with payment systems (which only generate any revenue if I spend to advertise). 

I maintain my own website, http://www.dickpountain.co.uk hosted for free on Google Sites and very plain in appearance and features, though I did recently port it to their New Sites standard. It contains a few short essays on computing, music and politics but largely exists as a hub from which to access all those 20+ other repositories. My SEO skills are fairly modest so people are probably more likely to arrive at one of those by Googling rather than via my site, and when they do arrive they’re unlikely to cough up any cash because I spend nothing on promotion. Making a living online has never figured among my life goals, but I do like people to see my work.

The one online medium I’d never tried was the podcast, partly because I don’t really like my flat, East Midlands speaking voice very much when recorded. However I did recently participate in one, via an extremely circuitous tour of the contemporary media landscape. In 1990 I drove to Prague with my brother-in-law Pip in his vintage 1937 Lagonda car, our purpose being to see Vaclav Havel installed as president and to witness Pip’s friend Berty – who’d had to flee the Russian invasion in 1968 – be given the keys to the city. Fast forward to 2017 when an old friend and colleague Mark Williams started a magazine called Classic Motoring Review and asked us both to write up our trip for him. Skip forward again to October 2019, Mark’s magazine has sadly folded, but I post our article from it to Facebook. Flashback: in 1983 Pip had founded the Scotch Malt Whisky Society to spread knowledge of the virtues of unfiltered, cask-strength single malts. Forward again to Oct 2023, Pip shows my Facebook post to the editor of the society’s magazine Unfiltered who decides to republish a full-colour version with an attached podcast of us reading it. 

Now Pip lives in Montrose, I live in London and the society is in Edinburgh so meeting to record was out of the question. I scrambled to test Android audio editing apps, found two that worked (Lexis Audio Editor and Bandlab), recorded my half and you can judge the result here https://unfiltered.smws.com/unfiltered-01-2024/smws-adventures-prague.While it was quite satisfying to add voice to my media types, it did prompt a rather morbid thought. The world’s in a febrile and unstable state right now (to put it mildly) so how long can one expect one’s cloud content to survive, say after a catastrophic cyber-attack like the one in the recent movie ‘Leave The World Behind’. That could wipe the lot. Once upon a time when an author died, people went through his or her papers, the books on their bookshelf, visited their publisher. All solid, material stuff – paper, cardboard, photo prints, paintings, film, tape. How’s that going to work in cloud land? Once Google goes down it’s all just so much scattered data. 

[ Dick Pountain needed a dram of Caol Ila after he finished this column]  

Friday, 12 July 2024

VINTAGE

Dick Pountain /Idealog 354/ 06 Jan 2024 10:04


For 30+ years now I’ve been dabbling in the digital arts in music, in graphics and indeed  programming itself. Digital offers enormous advantages in speed of production, editability and economy of experiment – no wasting paint and canvas, no twanging of strings or blowing on hard reeds. But for several years I’ve felt an increasing pull back toward the physical, and to judge by the content on YouTube I’m by not alone. 

If forced to change professions I’d probably want to become a luthier, repairing and fettling fine guitars (I wasn’t good enough at school woodwork to actually make them). My first job would be to reset the neck angle on my beautiful red 1963 Hofner Verithin, currently unplayable, which would make it worth around £1000. But the money is only part of the pull  I, and many others, feel towards restoring vintage objects: there’s profound satisfaction in the process and accomplishment too. Old cars, motorbikes, musical instruments, toys and more can become more valuable (sometimes staggeringly so) with age. Most computers on the other hand become fit only for the skip.

OK there’s a small community who cleverly deploy the marvellous little Raspberry Pi board to emulate and hence resurrect a handful of desirable vintage computers like the Sinclair Spectrum or Apple Mac 1, but these merely illustrate the problem. The speed of progress in the IT industry makes its past vanish. TTL chips that are no longer made, lost storage media (8” floppy sir, you’ll be lucky), operating systems, applications but most of all vanished esoteric knowledge. People who restore old cars or motorbikes can club together to have spares made by small metal-work shops. People who restore violins or guitars can do it themselves with planes and chisels. But fabricating lost chips or etching PCBs to restore an old computer isn’t feasible, and is also pointless if you can emulate them so effectively in software.   

Wearing my other, social science, hat I’m a fan of the French sociologist Luc Boltanski who has studied the way that objects acquire value. In a 2014 paper ‘The Economic Life of Things’ he and Arnaud Esquerre outlined three-stages in the ageing of manufactured goods, from milk bottles to motor scooters. In the first stage objects get used and their price is determined by the cost of materials, energy and labour needed to make them; in the second stage people start to collect certain used objects – beer-mats, stamps, bayonets, whatever – and value them by rarity. The third stage, the asset, sees the most rare and sought-after objects become too valuable to use – think Swiss watches, old master paintings, Ferraris:  they’re purchased by the rich as stores of value, thought to be safer than money or shares. Boltanski and Esquerre analyse these stages along axes like disposability, uniqueness and portability and though they did include laptops and smartphones, it was only in the first stage.

However 2014 is a decade ago now, a long time in IT, and that decade saw YouTube rise from merely distributing amateur videos to a source of substantial income for lots of people, by making videos about precisely what I’m talking about here. Videos about restoring everything from coffee grinders to aeroplanes, do-it-yourself videos from metal-work and carpentry to dangerous chemistry, and reviews of all kinds of objects. Many of the objects involved belong in the collectable class, while some, like Gilmour, Clapton or Cobain’s guitars, are £1m+ assets. Comparatively few videos involve restoring computers, maybe because soldering is far less photogenic than carving or sand-blasting or blowing things up. 

I’ve spent a frightening amount of time watching YouTube videos in recent years. My favourites include ChemicalForce, a man with an almost comic Russian accent who obtains dangerous chemicals, mixes them and choreographs the resulting explosions in slow motion and gorgeous colour with balletic skill; Hand Tool Rescue (stop that sniggering) who discovers and restores marvellous 19th and early 20th-century engineering tools; and any number of blacksmiths who bash wrought iron into beautiful utensils (my father worked in the Sheffield steel industry). 

I also sporadically watch The Trogly’s Guitar Show, a daily vlog-style channel by a nerdy Texan who makes big bucks dealing online in modern Gibson and Fender guitars, buys and sells from other online sites, thinks nothing of splurging $5000 on a promising one. My favourite of all though is a genial, bearded Canadian chap called Ted Woodford who maintains and restores vintage guitars. After watching him do several dozen neck resets I feel, or perhaps delude myself, that I could do one myself on my Hofner. Virtual worlds and exotic AI-generated scenes are beginning to feel thin compared to the real chisels, spanners and explosions on YouTube, even though both only happen on my screen. 

[ Dick Pountain fancies himself as a blacksmith as well as a luthier]

Tuesday, 28 May 2024

RENTAL HYGIENE

Dick Pountain /Idealog 353/ 06 Dec 2023 12:02


My ideas for this column often come from newspaper stories, and this one was about Meta (The Corporation That Used To Be Called Facebook, or MTCTUTBCF for short) which has just launched ad-free, subscription-only versions of Facebook and Instagram to comply with a new EU privacy law. That law, first deployed in Norway, prohibits tech companies from forcing users to agree to targeted advertising to use their products: the user must be free “to individually refuse consent to certain data processing operations … without being forced to completely waive the use of the service.” 

Charging money for the product instead of monetising users’ data through advertising is one way to comply, hence Meta responded by offering paid subscription versions of Instagram and Facebook as “a valid form of consent for an ads-funded service”. This means all the major online platforms now offer a choice between a ‘free’ ad-funded and a subscription or rental version: Amazon has Prime, Google has YouTube Premium, while content-providers like Netflix and Disney+ offer no other model. 

Economists of both left and right persuasion tend to frown upon ‘rent-seeking’ – that is, making money from what you merely own rather than what you make or what you do – as it’s reputed to stifle innovation and competition and is often a symptom of monopoly. That’s as may be, but there are obvious cases (for example water, electricity and communications utilities) where renting their service is the only sensible route – Margaret Thatcher’s attempt to get us to own a slice of them fizzled within months as people sold their windfall shares to buy wide-screen tellies.

Personally I have very few software services, like Spotify, that I use often enough that subscription makes sense, but I do rent 100Gb of extra cloud storage from Google for a very reasonable £20 per annum. What has me seriously irritated though is the recent wave of small and far-from-essential vendors rushing into the rental-only model, presumably as a survival tactic in the face of dwindling revenues. 

I recently went shopping for several music-related Android apps, including a guitar chord guide, a DAW (Digital Audio Workstations) and a video editor, and was astonished to find that so many of the competent ones now employ the same come-on, a ‘free’ 7-day trial but give us your credit card number now (tantamount to inertia selling, the weapon of encyclopaedia salesmen through the ages). I did eventually find one of each that was good enough and offered an old-fashioned “give us £XX to remove the ads and own it” deal with which I’m perfectly happy.  

A similar Gadarene rush into rent-seeking appears to have infected many of the content sources I’ve seen relying on for years, who have recently started teasing with the first few paras of an article before greying-out the rest and demanding my wallet. I’d often be prepared to pay a reasonable sum for a single article (though many academic sites want far-from-reasonable bucks, assuming your institution is paying) but I really don’t need yet another monthly subscription.

I’ll confess that over the years I’d cultivated an Artful-Dodger-like skill in getting over, around or under paywalls: once it was enough just to Google the article title outside the site (they got wise to that), then more exotic tricks like reading the Pocket-saved version (they got wise to that too). Now I just don’t bother with them as I have too much to read anyway, and I suspect many others feel the same way. 

This is not a traditional kind of rent seeking which exploits monopoly over an essential service, but instead vendors manipulating the sheer ease of the buy-now, one-button-press billing/payment process to convert customers to ‘subscriptions’ that they neither want nor need. It’s always been a problem for software vendors that their products are expensive to develop, cheap to copy and become out-of-date quickly and rental can look like a solution to all those problems, but it really does diminish the incentive to innovate or improve features: shrink to a skeleton team of maintenance developers, move to a subscription model, hey presto all revenue becomes almost pure profit, which keeps investors happy and pressures your competitors toward a rent-seeking model too.

Because people can be persuaded that digital goods aren’t quite like physical goods, renting them may weaken ownership rights over a purchase that can be turned off if you don’t pay up. This is what has the EU worried, not just about the services themselves but about the users’ private data – invalidating ownership rights without the explicit consent of the owner is tantamount to theft. Call me old-fashioned but I prefer to own the tools of my trade. I carry out pretty frequent audits of my subscriptions, of how much I use them, and dump without mercy. 


[Dick Pountain finds the ads on Facebook more hilarious than most of the content]

 




MY OTHER COMPUTER IS A LENOVO?

Dick Pountain /Idealog 352/ 06 Nov 2023 12:56


I’ve had a vague intuition that Android might become the number one OS in the world – simply because of its ubiquity on smartphones – but I wasn’t prepared for the actual market shares recently announced: Android 42%, Windows 28%, iOS 17%, macOS 7%, ChromeOS 1.3% and desktop Linux 1.2% (as of April 2023). 

This being so I wasn’t too surprised to read last week that Lenovo is teaming up with Android specialist Esper to revamp its ThinkCentre M70 line into an all-in-one desktop Android PC, using Intel Core processors all the way from an entry-level i3 up to i9s. I’ve made no secret in this column that nowadays I use a Chromebook rather than a Windows PC, but what I’ve not emphasised is that in effect it’s an Android laptop. ChromeOS offers a pleasant enough basic cloud file system, but ChromeOS apps are largely rubbish and its Web Store is a joke. Apart from Google Docs, in which I’m writing this, everything I do is done via Android apps. Strictly speaking of course Android runs inside ChromeOS on top of Linux, but I rarely go there, so let’s not.

It all works remarkably well and the common fallacy that there aren’t any grown-up Android apps is just not true (though admittedly it’s taken me a while to find them). Apart from Docs I now have an excellent photo-editor, draw/paint apps, MIDI editor and DAW that serve very well in place of Office, Photoshop and Ableton. Freedom from Windows upgrades and malware horrors is delightful, and backup ceases to be a nightmare, except…Except, I’m so paranoid by temperament that although everything is ‘safe’ in Google’s cloud, I want it local too. Hence I keep a tiny 128Gb USB stickette permanently occupying one port and write all my data – not apps – to that. (A couple of times a year I also copy this stick to a separate hard-drive for archive). Google wouldn’t approve, but it keeps me happy. This does somewhat restrict my choice of Android apps to only those grown-up enough to recognise external drives, and it complicates my mental image too because now there are three separate memory spaces to consider: the Cloud, the Chromebook’s internal memory and the USB. Such triplicity turned around to bite my hand, albeit mildly, recently. I’d started getting messages that the computer was running low on space, and the Files app did indeed show that only around 30Mb of the 24Gb internal memory were free. It turns out, unsurprisingly, that most of the space was being eaten up by Android apps, which appear under a tree called Play Files (named after the Android Play Store). Within that tree 6.5Gb of the space was being consumed by the subfolder Android/data, and when I peeked in there I felt a familiar creeping spine-shiver – familiar because it reminded me of years of trauma in the dank, dripping Windows Registry. It seems that most Android apps you install create their own subtree there, but few remove it fully, or indeed at all, when you uninstall them. And it was then that I discovered you can’t delete folders in Play Files using the Files app, because you lack the Linux permissions.  

After an exciting couple of days where I saw ChromeOS naked, not a pretty sight, an alternative file manager called Cx File Explorer told me that Play Files is really called system/storage/emulated/0/ and the USB drive is called system/storage/7875C92D409CB3B4F21633193CC4E1DFSAE2FAB7 (system programmers are a curious breed, a bit like that bloke in The Hurt Locker). It also let me delete most of the cruft, with much pondering over which might still be important. One alone, DCIM, the digital camera folder, still resisted deletion and reported its size as 635Gb, clearly absurd. Eventually I nuked that one from orbit using the Linux console and nothing broke: photos now arrive in Downloads just fine and I can keep a workable 6Gb of free internal memory.

The moral of this story? I still like Android a lot, it’s an excellent cushion-cover for the prickly horror that is Linux, and I’ve even acquired the ability to write GUI apps in Python for it. What’s more I’m not ashamed of my belt-and-braces affection for local storage and hope that perhaps Lenovo’s partnership with Esper might prod Google into completing the proper integration of ChromeOS with Android that it promised back in 2015 but never delivered. Given Microsoft’s never-ending nightmares with post-10 Windows versions, and given a whole generation reared on Android phones, Lenovo could be onto a winner if it dares go beyond its stated targets of the retail, hospitality and healthcare industries and tests whether keenly-priced Android PCs could become giant killers. 


 [Dick Pountain sometimes feels his own internal memory is down to 30Mb]

FAKING FRIENDSHIP

Dick Pountain /Idealog 351/ 07 Oct 2023 02:08


“Man told to kill the Queen by his AI girlfriend!” History may some day recognise that as the most perfectly-formed tabloid headline ever, but right now it’s merely a symptom.To put it another way, it’s an answer to the question ”what do you get if you cross a moral panic with a feeding frenzy?” I’ll have to admit that I hesitated before embarking on yet another column on AI, which I’ve been covering for over 40 years, right from the days of Lisp, Eliza and ‘expert systems’ up to creating pictures of yucky seafood dishes with Stable Diffusion. A couple of events forced my hand though, convincing me that evading the topic isn’t an option this month.   

The first of those events was the CogX Festival 2023. I first covered this event back in 2019 when it was held near where I live, at the newly-opened King’s Cross Granary Square site. Around 10,000 people attended some 500+ sessions given by leading AI techies from the UK and USA. The audience was overwhelmingly young, 20-somethings wearing backpacks and plaid shirts and either already working as AI developers, or wanting to. I was given a press pass and heard genuinely revelatory talks about deep-learning convolutions and IPU (Intelligence Processing Unit) architectures. 

During the Covid years CogX continued via Zoom and was inevitably far less exciting, but I still ‘attended’ perhaps a dozen sessions. CogX 2023 has just been held in the 02 stadium and attended by around 90,000, hearing over 1000 speakers on 10 stages, including celebrities like Stephen Fry and the Queen of Jordan. I didn’t even bother applying for a press pass and the ticketing structure was quite byzantine - you could get a free pass to get in, but then had to purchase ‘add-ons’ to actually hear anything, which could amount to anywhere from £500 to over £1000. I’d guess that’s simply because the emphasis is now on attracting investors rather than recruiting plaid shirts: AI has arrived, applications and commercialisation take precedence over research and development. 

The other significant event took place at home where, among several old friends we had to dinner were a director of an important science picture library and a professional photographer for a large media corporation. Both were apprehensive about the effects of AI on their businesses. The main AI large-language models have already ‘scraped’ the entire contents of R’s picture library, and there’s currently nothing they can do to prevent this. J travels the world photographing hard-to-capture events: a picture of his that once went viral was of a huge wave hitting a lighthouse in which the face of Poseidon appeared among the foam. He pointed out that nowadays anyone can replicate such a picture in Midjourney in five minutes without leaving their sofa. 

CogX to its credit still records the festival sessions and releases the videos for free on YouTube after it’s finished, so I went through the list of videos looking for in-depth technical ones that might spark a  future column. I found very few of those, but did find one that was highly relevant to that question I posed above – Gillian Tett of the Financial Times interviewed Yuval Noah Harari, author of the best-selling popular philosophy/anthropology book ‘Sapiens’.

Harari’s take on our current situation is that the first wave of social media sought to capture our attention by using clicks and likes, and sold these on to advertisers at colossal profit. But this will soon be displaced by a coming second wave of AI-driven media that instead solicit our affection by mimicking human feelings. AI agents will in effect give back attention to the alienated, isolated millions who believe that no-one is listening to them. This is a far more dangerous matter – instead of encouraging passive scrolling and shopping, it may recruit people to perform actions (like killing the Queen). Harari argues that AI needs to be regulated, sensibly but avoiding a total ban that would throw away all the potential benefits, and to this end he suggested to Tett three concrete policy proposals:

 1) To gain public trust, AI regulation will require international institutions that will have to be financed by a 20% levy on tech companies profits.         

 2) There should be an absolute ban on passing off AI agents as human beings (with jail penalties like those for counterfeiting or fraud).

 3) Bots must not be granted freedom of speech, and their owners should remain legally responsible for all their utterances.

I’d probably add to these that royalties be paid for content scraped from existing sources, similar to those from music streaming services (which themselves urgently need reforming). And perhaps some kind of vetting of AI girlfriends for homicidal impulses might be a good idea too.   


[Dick Pountain was disturbed to find that ChatGPT has better musical taste than many of his friends]

 

 


Monday, 25 March 2024

SOCIAL UNEASE

Dick Pountain /Idealog 350/ 07 Sep 2023 10:58 

 Ten years ago this column might have listed a handful of online apps that assist my everyday work, commenting on their effectiveness, but I can no longer do that – the plethora of apps and social media platforms have so merged, interweaved and entwined that they’ve now become my workplace. I spend much of each day in front of a screen reading, writing, communicating, drawing, coding, listening, the various apps coalesced into a complete world more real, and far more useful, than virtual playpens like Zuck’s Metaverse. 

It goes without saying that I’m writing this in Google Docs, and when finished will send it to Tim with a couple of keystrokes. I still prefer to communicate by email even though most of my friends have defected to WhatsApp, mostly for the free video calls. I detest it (as I also do Instagram) because I’m still laptop- rather than phone-oriented so their UIs don’t really work well. I do get out of the house occasionally for a walk, and take a few phone pictures. When I get home they’ll be there on Google Photos and I’ll edit one in Snapseed (or Artflow or Sketchbook, using Share to flip between them). I mostly post the result on Facebook nowadays, more rarely on Flickr because posting there feels increasingly tedious. I have a paid-for Premium account on Spotify, mostly to research new music, but I consume more music on YouTube when good videos are available. Any fantasies I once entertained about making money online have evaporated. 

To get monetised on YouTube nowadays you need a home studio, proper lights, and to cultivate winning/irritating mannerisms, none of which I’m prepared to do. But in any case, getting rich online nowadays means TikTok, OnlyFans, Substack – not YT – and being 20-something and wearing false eyelashes, none of which I’m prepared to do either. I’ve given up writing blogs (too mean to pay to promote them) and rejected podcasting as it requires a greater frequency of new content than I can muster. I do however spend/waste hours reading podcasts by those more motivated, like Andrew Hickey’s monumental ‘A History of Rock and Roll in 500 Songs’. By and large I don’t pay for these, Patreon notwithstanding. I also do enormous amounts of reading online, almost always in PDF or Kindle format, from platforms like Medium and the Atlantic magazine as well as academic papers. 

Articles that might be useful I print-to-PDF and store in my own (local) database. At this point it would be traditional to enter into the debate about whether this vast and spreading digital ecosystem will eventually completely supplant the older media, like paper newspapers and books, cinema, live music concerts. I’ll spare you most of that, to merely say that there are many omens that suggest otherwise. The original premise of this ecosystem was that internet access would democratise media production, allowing anyone a shot at online fame (remember an age ago when this was called Web 2.0, tee hee?) It sort of did though the result has been an explosion of volume with an implosion of quality. The amount of clickbait content on YouTube makes it a real chore to use. Digital streaming giants like Netflix, Amazon and co. are all in financial trouble because of the expense of generating endless streams of ever-more witless content, and because the cost-of-living crisis precipitated by Covid forces people to stop subscribing to stuff. Meanwhile the sheer hubris of macromoguls Musk and Zuckerberg tempts them to self-destruct their own platforms (though that cage fight would be worth watching).
 
Unsettling enough, but there’s worse to come, from several directions. The telcos are grumbling about having to invest in ever more bandwidth to feed the streaming moguls audience, without getting a fair share of the profits: 16 European telcos last year signed a joint statement calling for Google, Meta and Microsoft to pay more, since they have to spend €50bn (£44.5bn) annually building and maintaining full-fibre broadband and 5G networks. Then there’s the increasing pressure by governments to regulate online content, lead by the EU’s Digital Services and Digital Markets Acts which seek to curb online hate speech, child sexual abuse and disinformation by law, and end Silicon Valley’s status as its own gamekeeper. Even the UK’s Online Safety Bill might add a face-slap to the EU’s arse-kick. And then there’s the Hollywood writers’ strike, which threatens to go beyond matters of current payment and become an all-out offensive against the possible future deployment of AI content generation by the studios and streamers. And I still have paper subscriptions to the London and New York reviews of books, partly to support the future of long-form journalism, partly because they’re easier to read in bed over coffee and toast…

 [Dick Pountain likes to kid himself that he’s an analog influencer]

PORTRAIT OR LANDSCAPE?

Dick Pountain /Idealog 349/ 04 Aug 2023 10:25 

 I don’t know whether it’s just me becoming a grouchy old man, but TV adverts seem to have become more imbecilic over the last year or two, and one that particularly saddens me is when that girl sees someone using a Samsung Flip phone and is so overcome with consumer-lust that she runs away to join ‘the flip side’. Esteemed editor Tim recently asked as a ‘masthead question’ whether we editors were tempted by such bendy phones and what would make us buy one, and I replied “The Galaxy Fold tempts me for use as a camera, and what would make me buy one is a price reduction of exactly one order of magnitude”. The reason I fancied the Fold (price apart) is simply that it alters the screen’s aspect ratio from portrait to almost square. 

The shift from desktop PCs to laptops barely affected the aspect ratio via which we absorb our digital content, which remained mildly landscape at around 4:3 (also roughly the shape of cinema and TV screens in olden times). Cinema went widescreen from the 1960s but it took till the early 2000s for laptops and TVs to follow suit with 16:9 or 16:10, still landscape but more so. That all changed at a stroke when the smartphone took over the world, with its thin upright portrait format around twice as high as wide. (And before you say it, I do know you can turn a phone on its side, but people only do that to watch movies and the user interface is designed for portrait). Does this matter? Well, yes, to me it does, but I only barely understand why it is that I prefer landscape. Is the real world portrait or landscape? That depends on where you live: the highlands of Scotland are awesomely landscape but where I live in central London is very much portrait, streets lined with buildings that obscure any horizon, some so tall you have to lean back to view them. 

 The very terms themselves of course evolved in the era of classic oil painting, when rich folk who could afford it had one done sprawled with dogs and dead pheasants at their rolling country estate, and one upright in a chair in the town house. Portrait puts you at the centre of attention, as millions of selfie-sticks will testify. As a keen photographer I’m sensitive to the effects of field–of-view on composition, of different focal lengths and of film formats (the real pros often prefer square). What about physiology? The human retina is circular (more exactly a section of sphere) but the eye itself is oval thanks to its lids, with a horizontal field-of-view around 178o and vertical of 135o (60o upward and 75o downward). When looking straight ahead your eyes have a 95o field of view from nose to periphery, so rather unsurprisingly our vision leans toward the landscape, where we had to chase things to eat for several hundred thousand years. But I no longer have to chase my dinner over the hills, so that can’t be it. More to the point is that my digital existence nowadays lies in the borderland between phone and laptop. 

I use a Chromebook that looks like a landscape laptop but also runs Android, so that many of the apps I try were designed for phones with a portrait UI. The better ones do let me expand the window to fit and don’t assume I have a touch screen: the crappier ones do neither and go straight to bin. Among the big guns Facebook drifts inexorably toward portrait, following its acquisition of Instagram and introduction of Reels, while Amazon, because of Kindle, has become skilled at adapting to different screen and font sizes. Nowadays I read so many books in Kindle format that I’ve started using my Galaxy tablet again, and the non-fiction and tech papers I read or review are mostly PDFs that adapt easily to any aspect ratio. For many early years I was Dennis Publishing’s print buyer and hence intimately familiar with paper sizes. 

The printers’ A sizes all have the same aspect ratio of √2:1 or 1.4142, so that folding them in half - when printing book or magazine sections - retains the same aspect ratio. I’ve often wished, in vain, that the computer industry had adopted these A sizes so that turning a phone or tablet on its side could neatly flip into two pages. That would have made phone screens a bit fatter, and PC/laptop screens a bit taller, and more like paper books. So my preference for landscape is probably, as one who’s been addicted to reading since childhood, for the comfort of double-spreads in an open book or magazine…

 [Dick Pountain has an aspect ratio of roughly 4.6:1 (when standing and viewed from the front)]

WAVES RULE?

Dick Pountain /Idealog 348/ 05 Jul 2023 04:19 

 There appears to be a widespread opinion that biology and maths don’t mix. Perhaps it’s behind the current panic over AI, and the depiction of robots as villains in superhero movies (bring back the more sociable Robbie?) It was already prevalent in the 1960s at school when I had to choose which A-levels to take: to do chemistry I was told firmly I must take maths and physics but not biology or art, both of which I loved. In those days, and at university, and long afterwards, the only real connection between maths and biology was via statistics which were and remain necessary for designing and interpreting experiments and stuff like population studies. My interest in biology didn’t go away (I ended up in biochemistry) but it combined with interests in philosophy and computation to lead me down ‘eccentric’ paths. 

I discovered D’Arcy Thompson’s magnificent ‘On Growth And Form’ which showed how geometry was expressed in the shapes of living things. I encountered the Belousov–Zhabotinsky reaction where a certain chemical mixture oscillates between different states rather than proceeding smoothly to a final product: the Belgian chemist Ilya Prigogine (no relation) received the 1977 Nobel prize for extending this insight to explain what he called ‘dissipative systems’ which can organise themselves to exhibit complex moving structures.Last week in the Imperial College alumni magazine I read an article about professors Robert Endres and Mark Isalan whose work with synthetic embryos is revisiting the patterns described by Alan Turing in his seminal 1951 paper ‘The Chemical Basis of Morphogenesis’. 

Turing had suggested that biological shapes and structures might arise when two different chemical compounds diffuse into one another, and he proposed a mathematical description using partial differential equations, whereby waves of chemical composition flow through the growing organism catalysing reactions that create structures and control their placement in space. Though Turing pioneered modern computing, he didn’t yet possess one powerful enough to solve these equations, which he did by hand using approximations. Turing was also working before Watson and Crick discovered the structure of DNA, and the revolution in molecular biology and genetics they triggered now lets us understand the mind-boggling chemical systems that support his proposal. 

The embryos of multicellular creatures like ourselves start out as sheets of undifferentiated cells, but genes within some of them start to exude proteins called transcription factors which travel across the sheet in synchronised waves, turning other genes in far cells on or off and causing them to divide, to die, or to migrate, and to release further waves. This fantastically choreographed cellular ballet must happen in time and space very precisely to sculpt the shape of each particular creature: similar small sets of so-called homeobox genes generate the wings of a fly, the leaves of a tree, the segments of a worm or your arms and legs, via waves of proteins intersecting and interfering at the right places and times. 

 If waves intersecting and interfering to make stuff happen sounds familiar, that might be because particle physics, which of course underlies the whole of modern chemistry, is currently based on waves too. The Standard Model, which has so far survived all attempts to surpass it, proposes that the universe consists of nothing but a set of ‘quantum fields’, one for each kind of particle, currently 17. All matter and energy and things and people, everything that happens (except for gravity at the moment, to the great chagrin of physicists) happens via ‘perturbations’ of these fields which spread like waves through space and time. And now for the punchline. Last week NANOGrav (North American Nanohertz Observatory for Gravitational Waves) and various international collaborating observatories released results that suggest there’s a very low frequency background ‘hum’ of gravitational waves permeating the whole universe. This hum might be the result of supermassive black holes merging, or it might be left-over ripples in space-time from shortly after the big bang, and it’s only now become detectable by new more precise LIGO (Laser Interferometer Gravitational-Wave Observatory) arrays that exploit ultra-tiny changes in the frequency of pulsars. 

But already there’s speculation that at some time shortly after the big bang, before there was light, just a soup of newly forming particles, it could have been interference between waves in this all-permeating gravity wave background that caused them to clump together into the first stars and then the first galaxies. There’s an apocryphal story about Bertrand Russell giving an astronomy lecture where a woman in the audience challenged him by saying the earth is actually supported on the back of a giant turtle. When Russell asked what that turtle was standing on, she replied that it’s “turtles all the way down.” Perhaps she was almost right but it’s the waves rather than the turtles… 

 [Dick Pountain would rather waive the rules than rule the waves]

SEARCH ME

Dick Pountain /Idealog 347/ 06 Jun 2023 11:22 

 “I'm writing this column sitting under a magnificent copper beech tree in the West Meadow at the Kenwood Estate in North London, with tall summer grass all around and even a foxglove in the corner of my vision. You might not find this any big deal - after all mobile computers have been around for a long time - but it has taken me personally twenty years to achieve this feat, and I've only achieved it now thanks to my Palm Pilot and a neat little add-on keyboard called the GoType, whose purchase was the best £70 I've spent in a long time.” 

 In actual fact I’m writing this column sitting under a magnificent copper beech tree in The Regent’s Park 24 years later, on a Moto g8 PowerLite phone, and I’ve just pasted that quote from Idealog 59 (1999) in seconds merely by searching for ‘copper beech’. That was possible because I post my old columns, six months in arrears, onto Google’s Blogger, which has an excellent full-text search facility. Of course this search only works on plain text files, not binary or encrypted ones, so even though over the years I’ve written columns in TextPad, MS Word, Libre Writer and now Google Docs, I upload them by cut-and-paste as plain text. 

The searchability of archives bothers me a lot, because I gave up on huge document management systems and NASes long ago in favour of simplicity, and that’s not an easy row to hoe. Recently an old friend, who’s just retired as a professor of robotics, mailed me to ask for a Byte article I wrote about his company back in 1985 and I was keen to oblige. I found the right Byte issue on the Internet Archive using their own search engine, which gave us a PDF of the page, and I located the original copy file on my local machine too. But this experience set me to thinking that I really don’t have an effective full-text search on my Chromebook that can locate topics in the many decades of files I keep locally (I love The Cloud, but I don’t trust it). I was about to write that this wasn’t a problem when I was a Windows user, thanks to its indexing system, until I recalled the appalling problems I used to have with that system consuming vast amounts of CPU time, and occasionally going metastatic and filling my whole hard drive with its index file. 

 Anyway, I checked out grep-like apps in the Android Playstore only to discover they were all indescribably useless – several didn’t work at all, all were full of dodgy ads, and almost all failed to access external USB drives. I got so fed up installing and uninstalling these monstrosities that I formed a crazy idea that I needed to write my own. I entertained that idea for several minutes, before remembering that I’m working with Android running on Linux within a ChromeOS cocoon, and that I’ve always hated I/O programming at the best of times. So I had an even crazier idea, that I would ask my new friend ChatGPT to write one for me. I’ll admit that I considered this more as an experiment (and possible column topic!) than a serious strategy. So off I went: Me: “write python 3 program to find address of external SD card on Asus C301S Chromebook” ChatGPT: “On Chromebooks, external storage devices like SD cards are mounted under the /media directory. You can use the os module in Python to list the contents of the /media directory and find the address of the external SD card. 

Here's a Python 3 program that accomplishes this: … “ That Python code didn’t work quite immediately because it turns out the SD card mount directory name isn’t ‘/media’ on my machine but actually ‘/storage’. With that one alteration it just worked. I asked for a second Python 3 program to search for a target string within multiple text files, spliced them together with a couple of extra lines to format the output the way I wanted and it was done, in a couple of hours. My new app worked perfectly except for a rather amusing bug: some of the Byte columns would crash it with a Unicode error, which I discovered was because back in those prehistoric days my text editor used to insert a character code for the ‘£’ sign that conflicts with 64-bit Unicode. Python’s ‘try…except’ error handling soon put a stop to that. I was deeply impressed (and a little scared) by how fast ChatGPT generated such excellent code, but I don’t think I’ll be pushing my luck by asking it for an app that can search within DOCX and PDF files too. 

 [Searching Dick Pountain’s 341 columns at https://dickpountain-idealog.blogspot.com/ will reveal only one occurrence of ‘copper beech’]

TURNING THE AIR BLUE

Dick Pountain /Idealog 358/ 07 May 2024 01:32 In my back-room hardware morgue is a black cotton bag, about the size of Santa’s Sack, contain...