FRACT V FRICTION

Dick Pountain /Idealog 322/ 07 May 2021 01:10

There’s never been a better time to be a digital artist. Last night Channel 4 News had David Hockney talking about his new show of Spring landscapes painted on his iPad, at the Royal Academy. And back in March Michael Winkelman, aka ‘Beeple’, sold an NFT (non-fungible token) to his work , a collage of 5000 digital images, for a record $69,400,000. It was purchased by Vignesh Sundaresan, aka ‘MetaKovan’ who paid for it using 42,329 Ether digital currency.

I’m not quite in that class but I do consider myself a digital artist, and it must be true because I’ve put it on my business cards. It’s not just that I take digital photographs, which I discussed in last month’s column. I’ve also written here, a few years ago, about the way I took such photographs and turned them into ‘paintings’ by repeated transformation in Photoshop, using layers, filters and other effects. This kept me amused for quite a while, but eventually it palled: such processed photos have become a bit of a cliche in adverts and music vids, so I gave up for a while. 

Then in a major life change I deserted Windows for a Chromebook. I tried Photoshop Express online, but that didn’t cut it so I sought alternatives: there are capable photo editors like Pixlr, but none had all the filters I needed until I came across Sumo Paint, a sort of minor Photoshop clone for Chrome but with several whacky features of its own. One of those features was a ‘fractal morph’ filter, which takes any image created in the draw and paint section and applies a fractal transformation to it, with several adjustable parameters that create very striking, naturalistic looking images with fractal levels of detail. I soon became hooked, stopped working from photographs and started drawing simple, colourful, textured shapes, fractalising them, cropping and further processing them to create a new (to me) medium.

The results were pleasing, looking like very complex paintings that might have taken months to do with brush and pigment. And that after all is what digital art – from painting to synthesisers and drum machines – is all about, creative cheating, skipping the hard slog. You can see some of the results for yourself at my web site.

Greatly excited I naturally wanted to try and sell these images, to become a real artist who gets mashed in the Groucho Club every night. There were several obstacles though, the first and worst being resolution. Fractal morphing is very compute intensive, so runtime explodes with increasing image size and eventually crashes. The biggest initial drawings I could process were between 600x600 and 1000x1000 pixels (dependent on content), and though the results look great on screen, they’re too grainy when printed large. A partial solution is to resample them at 4000x4000 with interpolation, which looks smoother but fuzzes out fine details. A second obstacle is that no-one buys abstract prints like these off screen, and the substantial cost of printing and framing a lot of them on spec was an investment I was reluctant to make. 

A third obstacle is that Sumo Paint was Flash based, and Google terminated Flash support in Chrome in February 2021. Sumo’s authors brought out a new, non-Flash version – and it doesn’t have the fractal morph filter...  When I contacted them they told me they’d permit the older Flash-based app to run on Windows, so I now keep a digital museum piece: a Windows 8.1 installation running discontinued Flash and Sumo Paint. It feels rather precarious, a bit like painting in digital egg tempera.

Things have perked up a bit on the printing front though. A friend of mine who paints the old way in acrylics started putting her work on the Saatchiart website and selling plenty of it, and recommended I give it a try. Saatchi offers two different deals: they’ll print, frame, sell and deliver for you, deducting their commission, or they’ll pass buyers on to you and you deliver (which obviously she has to do with paintings). I’ve put some of my work up there under the first deal, but I’m getting nowhere near her level of sales because I’m selling only prints rather than ‘original artwork’, and most people who spend serious money on art want to own the original. Which makes me very curious indeed about Beeple and his NFTs. It seems crazy that people are prepared to spend millions to not own the original artwork, but it’s not that much crazier than Bitcoin or Ethereum, and when everyone is going crazy then it feels a bit crazy not to go crazy too.      

[Dick Pountain invites you to "get in on the ground floor" at https://www.saatchiart.com/account/artworks/1239765 ]

 

BE STILL MY HEART

Dick Pountain /Idealog 321/ 09 Apr 2021 01:58

I like to cook, and although I have two whole shelves-full of cookery books, digital sloth dictates that nowadays I tend to go online and Google for a recipe rather than trying to find one there. Googling ‘pork mushrooms kimchi’ will get me hundreds of matches. But once I click on a promising candidate, if it takes me to a video on YouTube I’ll back out sharply and try the next, until I find one that’s just a still photograph and a text recipe. 

I spend an inordinate amount of my free time watching Japanese sushi slicers, guitar neck resetters, jet fighters and restorations of rusty mangles on YouTube, so it strikes me as odd that I have such a powerful aversion to watching cooks performing a recipe that I want to do. Why is it? It’s only partly pragmatic, because most of these online cooks take so long to do stuff and to talk about it, compared to me just reading an ingredient list. And it’s only partly because so many of these cooks are intensely irritating (indeed a few of them are quite cute). No, it’s something more fundamental than that, and I felt a column coming on…

The difference between a still photograph and a video runs ontologically deep. A still photograph does something that’s otherwise impossible, namely it stops time, which your living eyes and brain refuse to do. That’s not all it does though: unless you’re a superb professional photographer, it’s your camera rather than you that decides most of what goes into a picture. Certainly you decided this was a scene or moment that you want to record and remember, but most likely it was only a couple of objects in it that sparked your interest, and much of the background detail will escape notice until later. Video is entirely different, behaving more or less exactly like your eye and sharing its inability to stop time: when you’re videoing you’re capturing a continuation of your ongoing perception. A still photo represents what’s strictly in the past, while video in effect makes past events present again. 

The great art photographers all took pictures in black and white mainly because colour film wasn’t invented or was so poor back then, but there’s a sense in which it was also right: the past ought to be monochrome because colour belongs to the living present. That’s why most kids under 30 find it hard work watching black and white movies, and why Hollywood is prepared to spend big bucks to colourise some of the classics.  

YouTube videos don’t just show us a resumed present but also their author, whose voice and ego are on display. A still photograph makes no comment apart from its manifest content: it’s the camera, not the author, saying ‘here’s the stuff that was in front of my lens when you pressed the button’. This quality of impartial historical commentary becomes increasingly desirable as our world becomes more and more swamped by 3D-animated, CGI-fied, deep-faked video. Hence the excitement when a large cache of old photographs is discovered, as happened with Vivian Maier’s Chicago street photographs and similar finds in Aberdeen and Hackney. We somehow feel that such pictures are less-tainted testimonials to the past (not true of course because they represent just one person’s choice of what to record).

Anyway, perhaps why I prefer still recipes over YouTube videos is that I really don’t want someone else’s opinion or experience of a dish, simply its bare, unvarnished facts so I can interpret them in the way I want. And this preference extends to my taking of pictures too: I’ve never been interested in shooting video (though I do enjoy hacking together GIFs from sequences of still images). I’ve been taking still photos for quite a long time: I got my first proper camera, a Yashica TL-Electro back in 1972 and my first pocket camera, a Canon Dial (half-frame, with clever built-in clockwork motordrive) soon after. 

Over the years I’ve been through quite a few pocket cameras - a Sony WX350 being the latest - but I’ve never succumbed to shooting with a smartphone until a couple of weeks ago. I’d clung on to an old HTC Desire phone with a crappy camera for as long as I could, but last week I couldn’t, so I bought me a new Moto Power. The quality of its cameras and flash impressed me, so off I trot to the park to take some pix of the Spring blossom. Back home the pictures looked odd and behaved even more oddly, until I realised that they were all actually videos. Those tiny icons for still and video are right next to one another, and in bright sunlight without my specs on… 

[Dick Pountain does not possess a selfie stick]

 

LET ME IN

Dick Pountain /Idealog 320/ 05 Mar 2021 10:07

Last month I confessed to feeling less hostile toward the giant digital corporations than is fashionably required nowadays, and this month I’m heading further still into the wilderness by admitting that I’m not particularly paranoid about online security either. That’s not to say I deny the importance of my dear colleague Davey Wilder’s advice which I do follow, and as a result have only experienced two successful virus attacks in 40 years. 

In the mid-80s I used to visit CeBit with Byte every year, where one drunken evening with members of the Kaos computer club I accepted a floppy disk of source code that blew up the laptop I was using. (Un)fortunately it wasn’t mine but borrowed from a friendly London dealer. Ooops. The second one I caught around 1998 after clicking a photograph on The Register which injected some malign entity that rapidly paralysed my own Thinkpad. Windows backup failed too and the plucky PC Pro lab folk had to dig me out of that one. 

Maybe it was that experience which persuaded me to sell my soul to Google. Since then Gmail’s antivirus abilities have stopped all manner of nasty attachments, and together with regular Windows Defender and Malwarebytes scans kept my laptop clean until I moved to a Chromebook. All fine, until recently Google’s vigilance has turned oppressive: I occasionally get a Critical Security Alert saying someone has just logged in using my password. I have four devices, two connecting via Android, one via Windows and one via ChromeOS and Google appears no longer to realise that they’re all me. I do change my password now and again, just in case it was true, and thrillingly Google refuses to accept my new password the first time I try it and asks for the old one, then accepts it and sends a Critical Alert saying someone has just changed my password. All very exciting, but small beer compared to some of the other sites I use.

I recently had a spat with PayPal when they upgraded me to a paid-for professional account I didn’t want. It was quite a performance trying to sort that one out as it’s virtually impossible to speak to a human and they place ‘limitations’ on your account that stop you changing anything. In the end after sending several emails from the help centre I was permitted to close the account and open a new, none-pro one that I now use satisfactorily.

But that again is a model of efficiency and approachability compared to Patient Access. A subsidiary of EMIS Health, this site provides appointment booking and repeat prescription services to GP practices, including mine. I only use one service, a 3-monthly repeat prescription for blood-pressure medication (not induced by Google or PayPal I hasten to add). This had been working well for over 10 years until last November, when Patient Access announced an extra level of security using a memorable word and hint mechanism. This didn’t bother me as my bank has been using a similar system for years, so I duly set both and forgot about them until last week when my prescription needed renewing.

Logged on with my old ID and password, was asked for three characters from the memorable word, which it said were invalid. Tried several times more, same, told only had three tries left. Decided it must have stored my memorable word wrong, so went to change it. Could only do that after receiving a code sent by SMS to my mobile, but since mobiles had never been involved before they had an old number from a previous phone. You guessed it, to update my mobile number I need to enter the memorable word….

I did eventually reach a human via email (after much digging to find an address) who told me there was nothing they could do except close the account, and I must re-register by visiting my GP surgery (pretty well inaccessible due to Covid). Fortunately I had linked the account to my local pharmacy years ago, which is now permitted to renew the prescription for me, so I’ve just walked away from the wreckage. Being curious though I Googled “Patient Access memorable word problem” and stumbled into a pit of boiling white-hot rage and indignation: hundreds were similarly locked out despite correct log-on details, some needing far more serious meds than me (one of them wondered whether the site had been designed by Dido Harding). 

In last month’s column I suggested that perhaps a good way to tax the digital giants would be by accepting part payment in use of their excellent infrastructures for our public services. Can’t imagine a better example of where Amazon’s seamless purchasing process is sorely needed. 

[Dick Pountain’s blood pressure is 128/76] 

   

  

TOO BIG TO LOSE

Dick Pountain /Idealog 319/ 11 Feb 2021 09:15 

Having used most of them since they first emerged, I’m far better disposed toward the giant digital corporations, aka Big Tech, than most people. I started working from home long before the current emergency, 38 years ago in fact, and doing that without Google and Wikipedia is unthinkable to me, while entertainment would be very dull without YouTube and Spotify. Amazon wasn’t quite so essential before it put Maplins out of business, but the last year has made it indispensable, while Facebook has become a place to gather, joke and grumble (often about FB itself). Twitter and Instagram I find irritating and only rarely use. 

This very long acquaintance means that I still regard Big Tech’s services as remarkable amplifiers of my abilities, rather than sinister threats to my freedom. I do of course acknowledge that their monopoly power, lack of care over fake news and hate speech and flagrant tax avoidance are very, very serious problems, but I believe they need to be cured rather than crippled or exterminated. I feel increasingly out-of-step in this though: more and more people I know boast of giving up one or more of these online services (usually FaceBook or Amazon) and claim to support drastic government action against Big Tech.

I’ve written here before about how their wealth makes the digital titans serious competitors to states, but I also remarked (as Joe Stalin once sneered about the Pope) that they don’t have armies. What they do have is a cosy relationship with the Democratic Party, and they do tend to be more progressive over climate change than fossil fuel corporations, so I was surprised how little action Donald Trump took against them (maybe because Twitter and Facebook were so essential to his rise: if he escapes impeachment to run again God help them). 

He did however initiate an antitrust lawsuit against Google which experts agree Biden is likely to continue, and several states including New York are filing their own suits to combine with the DoJ’s. Biden has hinted he might revoke Section 230 of the Communications Decency Act, which shields internet companies from liability for their content, while Congress Democrats have published a 449-page  on the monopolistic practices of Apple, Amazon, Facebook and Google. There’s talk of a national consumer privacy and data protection bill – like California’s Consumer Privacy Act  – which would restrict tech companies’ ability to store our data and thus severely damage their business models. It’s unlikely however that actual break-up will be on the cards any time soon. 

Many of the tech giants are ambitious to provide real-world as well as virtual services: Amazon, Google, Microsoft, Facebook and Uber are already working to penetrate healthcare, education and transport, using innovative, ‘disruptive’ services supported by their powerful infrastructures and AI capabilities which are often more efficient than state equivalents. They remain commercial, unelected, enterprises though, committed first of all to shareholders and whose incentive is to displace rather than create jobs. If they’re permitted to gradually take over services from shrinking states, promising to invest in carbon-reduction technologies while simultaneously creating mass unemployment, then states are tempted to lean on their technologies to integrate welfare, security and taxation, as is already happening with India’s Aadhaar system, Singapore’s ‘Smart Nation’ and Alibaba helping Chinese local governments to run the Social Credit system. There are serious questions about how effective such systems really are, but they certainly grant states a sinister degree of extra power over citizens, letting them punish dissidence by loss of benefits.

Political polarisation is weakening states throughout the Western world so that their very survival depends on delivering the goods more equitably: with the right policies Big Tech infrastructure could be made to help rather than destroy democracy. A spectrum of measures from outright break-up at one extreme, through increased regulation and fairer taxation, to public-private partnerships at the other (hopefully better designed than those New Labour finance-friendly initiatives that we’re still paying for) could enable governments to tame the corporations’ power without destroying them. Construct different ‘packages’ for different economic sectors – armaments, fossil fuels, pharmaceuticals, automotive, medicine, education –  by assessing the social value of each and choosing the appropriate set of measures from that spectrum. The precise composition would be negotiable, providing the leverage the state needs. 

The digital titans aren’t too big to be regulated, though ironically enough it’s the EU that has been most effective at doing it so far. Rather than simply collect more monetary tax from tech companies (who have made an artform out of evasion via haven countries) states might instead accept some of the payment due in free public use of their services, heavily discounted electric vehicles for public transport, or similar non-cash alternatives.

PICTURE WORTH HOW MANY WORDS?

Dick Pountain /Idealog 318/ 08 Jan 2021 10:58

Pictures mean as much to me as music does. I love paintings and used to regularly visit galleries before Covid closed them. The region of Italy where I lived for some years was within 40 miles of half the world’s greatest paintings and the birthplaces of Michelangelo, Piero della Francesca and Luca Signorelli. I don’t paint myself, having failed to bond with oils or watercolor, but I do love to play with digital images. (I could never consider writing a novel with pen or manual typewriter, and similarly I can’t imagine painting in any medium that doesn’t have Undo). 

For many years I’ve been creating fractal art using two main tools, Sumo Paint and Zen Brush. The latter emulates a Chinese/Japanese style bamboo calligraphy brush that you can use on a tablet with either finger or stylus: the original version was grayscale-only but they’ve recently released Zen Brush 3 which has gorgeous colour and very realistic watercolour effects. I do feebly try to sell work via the Saatchiart website (which has so far failed to make me rich) but at least they cost me very little in either time or money to make. The US abstract expressionist Philip Guston once observed that “The great thing about painting and drawing, as opposed to thinking about it, is the resistance of matter”, but it’s precisely that resistance that makes experimenting with canvas and oil paint so expensive: it confines lesser artists to garrets, and while overcoming that resistance makes a few of them great, I’ll just stick to dabbling on the cheap. 

Which brings me neatly back to one of the recurring themes of this column, namely that ’bits aren’t atoms’. I love to remind you that while you can order a pizza from a picture on your screen, you can’t eat the picture. That observation was actually true long before the computer age – you can admire a lobster in a Willem Kalf still life but you can’t eat it, and that painting was made using pigment particles suspended in oil, not bits. Representations of any kind, whether bits, or paint, or words on a page, aren’t the things they represent, even if modern technologies conspire to make us forget that. 

I’m a great admirer (wouldn’t say follower because he didn’t want to be followed) of the Spanish-American philosopher George Santayana. He’s not much remembered nowadays, and was never popular in England or the USA where linguistic philosophy still rules. 

His major mature work ‘Realms Of Being’ proposed that there are four such realms: Matter, which is all that exists and of which everything is made; Essence which consists of configurations, images and representations of matter; Truth, which contains just that subset of essences that actually correspond to material things; and Spirit, by which he meant the intelligence of living creatures, via which they perceive and process essences. We can never see matter directly, only perceive images of it, and that these images correspond well enough with actual material objects – so we don’t bump into trees or step off cliffs – is thanks to evolution honing our senses to fit our particular niche well enough. 

Most important of all, essences can’t do anything, they can’t affect matter directly: there is no magic. There are indeed dreams and imaginary objects, but they can only affect the world of matter if they persuade us to move, to do something. Santayana’s Doctrine Of Essences applies not only to the inedibility of digital pizzas, but to everything we do, and it’s a particularly powerful tool in these times of digital imagery, deepfakes and fake news. For example saying that we are ‘in control of the virus’ doesn’t affect the transmission rate of SARS-CoV-2 one little bit: only doing stuff, like vaccinating, hand washing, social distancing and mask-wearing can do that. Make America Great Again does nothing to increase the well being of Americans unless accompanied by policies that affect the material world. Sticks And Stones do indeed Break Bones, but words do not (though they can incite people to use baseball bats to accomplish that). A painting can change the world only by inspiring some people to do something, like go on a crusade or resist a dictator. 

In his recent book ‘Narrative Economics’ the US economist Robert J. Shiller attempts to apply a similar doctrine to the dismal science. He notices that people’s economic behaviour is not governed solely by self-interest nor by rational choice as current orthodoxies would have it, but by the stories people tell themselves, or are told by their friends or the media, about what is happening in the world and what that might do to their own future prospects. Take hoarding bog-roll for example...     

 [Dick Pountain is rather sorry that Augmented Reality can’t ever let you Undo the world]

GUITAR IN THE SPACE AGE

Dick Pountain /Idealog 316/ 06 Nov 2020 10:55

Music and electronics have gone hand-in-hand for me from the very beginning. The folk songs we were taught in junior school didn’t really grab me, and the first tunes that really did when I was 15 were two guitar instrumentals, Apache by Cliff Richard’s band the Shadows, and Ghost Riders In The Sky by US band The Ramrods. Both made gloriously excessive use of echo machines, and I badly wanted one.

 
Around that time (1960) Britain was still peppered with shops and dumps selling ex-WWII electronic surplus: we had a local one, a corrugated-iron Nissen Hut in a farmer’s field full of RAF-surplus radar sets and similar stuff. We used to hang out there, buying bits to make crystal sets - or even one-valve sets for the more ambitious - on which to listen to Radio Luxembourg, which was the only place, apart from café juke-boxes, we could hear proper rock ‘n roll in those days. I bought ex-RAF intercom carbon microphones, and attached one to a cheap, nasty Eko acoustic guitar as a pickup, to be played through the family radiogram as an amplifier. It sounded awful, but not awful enough so I made an echo machine using the coiled wire element from an old electric fire stretched between two more carbon mikes as a mechanical delay line. Then I learned Ghost Riders In The Sky and Apache.

A couple of years later I was among a team of 6th-formers who won a prize for building an analog computer that could solve sixth-order differential equations, using op-amps made out of RAF surplus components, including a green-screen radar oscilloscope as output. I went off to college in The Big Smoke where one could purchase real electric guitars (a Hofner Colorama) and real electronic effects, like a tape-based Watkins Copicat echo and the first Fuzz Face distortion box. I graduated from surf-rock to jazz, then to free jazz. Alongside an excellent bop alto sax player, I played gigs where I sat on a small Fenton Weill amplifier, playing my Colorama with lots of feedback and Mike Bloomfield-style bends and occasionally a violin bow. That amp had a crazy built-in tremolo effect, and by twiddling its rate and depth knobs I could make it work like a primitive beat-box.

I was never a jazz purist: in those days boundaries weren’t so sharp. I revered B.B. King and Hendrix and Clapton and Jeff Beck. I bought a Harmony Sovereign acoustic and learned ragtime blues finger-picking. I loved the classic jazz guitarists like Joe Pass, Herb Ellis and Barney Kessel, but didn’t want to play like them (clean-toned archtop guitars and fast fancy chord and scale work). Nor, later on, did I fancy becoming a heavy-metal shredder. No, tone was becoming most important to me, and electronics were the way to great tones.

Nowadays I have a nice vintage, Made-In-Japan Fender Stratocaster, an Ibanez Artcore semi-acoustic, plus looper, delay, tremolo and freeze pedals and a multi-effects box crammed with signal-processors that can emulate 100 different effects, amplifiers and rooms. This is a golden but whacky age for guitar playing, in which people debate endlessly about the tonal properties of different woods (and pay outrageous prices for them) but I don’t buy any of that: it’s all in the electronics.

Regular readers may recall my frequent mentions of the US guitarist Bill Frisell, which is because he’s been my hero and role model in all matters guitar for many years. He has pulled together all the genres that I most enjoy - surf rock, blues, pop, bluegrass and free jazz - into a seamless synthesis, but better still tone is his artform and effects his principal instrument. His pedal board isn’t large, it’s what he does with it: at centre is a Line 6 DL4 delay pedal which lets him not just record and repeat passages, but speed them up, reverse them and much more: stooped over to twiddle its knobs in mid-performance, Frisell gets unearthly sounds far deeper and richer than any synthesiser just by mangling the signal from his Telecaster.

During his long career Frisell has played with the best of the best, like ex-Miles Davis sidemen Paul Motian and Ron Carter, and has run two long-term trios of his own. If you want a taste of his music to see whether you’d like it, try his gorgeous solo rendition of the Beatles ‘Strawberry Fields Forever’ at https://youtu.be/DioXHQmGQko. If you liked that, a 1995 concert with two late greats, ex-Cream drummer Ginger Baker and ex-Ornette Coleman bassist Charlie Haden is much wilder: https://youtu.be/AwBtJm7JiDI while if you’re still on board after that, this marvellous set might hook you for good: https://www.youtube.com/playlist?list=PLwGTKZ6uNxXkrjkDlToJXU2EPG2awfGFx

A NOVEMBER TO REMEMBER

Dick Pountain /Idealog 317/ 04 Dec 2020 10:43

World-changing innovations are like London buses: you wait for ages and then three come along at once. The recent wait has been particularly irksome, as under pandemic lockdown virology and epidemiology felt like the only relevant sciences – apart from rocket science, to get us all off this pestilential planet and to Mars (joke). Then suddenly, blam, three arrive in the same month: I’m writing this at the beginning of December 2020.

Most important and welcome was the arrival of not one but three coronavirus vaccines, all produced in record-breaking time and apparently highly effective (though that will only become certain once they’ve been deployed widely for a while). While their speed of development, testing and approval is remarkable, that’s not the innovation I meant though. The innovation is that two of the three are totally synthetic. Most vaccines up till now have required the target virus to be cultured in animal cells, then deactivated or broken apart and bits injected into live animals to generate antibodies. The Pfizer and Moderna vaccines are for the first time purely chemical: strings of Messenger RNA assembled from scratch by those sophisticated gene-sequencing and nucleotide assembling machines that have been invented over recent decades by harnessing powerful computers to robotic chemical processing. Injected into your arm, this mRNA tells your own cells to start making one harmless spike-protein of the coronavirus, a sort of pretend infection that generates antibodies against the real one.

The second innovation is in the same area of molecular biology, namely a great breakthrough in solving the ‘protein-folding problem’. All the biological processes that animate living things are driven by enzymes, which are proteins - very long chains made up from 20 or so different amino acid units. These chains don’t just flap around like pieces of string but fold themselves into compact lumps whose exact 3D shape, their cavities and crevices, enable them to work by fitting the molecules they work upon in lock-and-key fashion. These lumps are held together by bonds between amino acids from different points in the chain, and to design artificial enzymes, or drugs that alter the action of natural enzymes by fitting their slots, then you must simulate the way any particular chain of amino acids will fold itself. Since every link in the chain can rotate freely this is a crushingly difficult computational task if the chain is thousands long.

Back in October 2002, PC-owning nerds were being encouraged to donate their spare CPU cycles to a world-wide distributed network for solving protein-folding problems run by Stanford University. 30,000 donors helped crack protein structures around 50 amino acids, but 100 amino acids would need another 270,000. On 30th November 2020 an article in Nature related how Google’s AI subsidiary DeepMind – famous for cracking the board game Go sufficiently to beat the human world-champion – had made a similar breakthrough in protein-folding simulation.

DeepMind’s program AlphaFold 2 had outperformed around 100 other teams in a biennial protein-structure prediction challenge called CASP (Critical Assessment of Structure Prediction). The performance of folding prediction programmes is measured against physical  experimental evidence of a protein’s structure from X-ray Diffraction Crystallography or the newer Cryo-Electron Microscopy, which pictures the individual atoms: AlphaFold 2 scored close to a 90% match in this challenge. An earlier version of AlphaFold worked entirely by deep-learning, that is by examining the amino acid sequences versus 3D structures of many proteins, which enabled matches of around 60-70% accuracy. For AlphaFold 2 the team added an extra level, not of deep-learning but of constraint-solving. Consider any pair of amino acids linked together and physical chemistry can tell you how they can rotate about that bond and what resistance will be encountered: similar data is available for the close approach of active groups on remote amino acids. Applying such constraints to a purely learned prediction can boost the accuracy to 90%. The implications for the study of human biochemistry, disease and the speeding of up future drug design are massive, potentially world-changing.

The third innovation unveiled in November relates tangentially to the other two: Apple released its M1 CPU chip, which has two perhaps world-changing virtues. Firstly it breaks Apple’s own dependency upon Intel and more importantly elevates the ARM architecture, which already owns the mobile market, to desktop PC status, thus threatening Intel’s x86 hold there. (Ironically enough our dimwitted government recently allowed ARM to be sold abroad). Secondly the M1 chipset contains not only eight general-purpose ARM cores and an 8-core graphics unit, but also a 16-core Neural Engine capable of performing up to 11 trillion deep-learning operations a second. Alongside those gene-sequencing and nucleotide assembling machines, M1-powered computers running AlphaFold-style software promise a new era of computer-aided biology, a sort of Lego with living cells.

[Dick Pountain would like an M1-powered Chromebook for Christmas (perhaps not this one)]  

BOOKISH MATTERS

Dick Pountain/Idealog 315/11:48 Monday, October 5th, 2020

One of my lockdown activities has been extending my reading of the great Italian writer Italo Calvino, which included his If On A Winter's Night A Traveller, a serio-comic, postmodern novel about books. The narrator (perhaps also the reader) begins a novel that ends abruptly because it's been wrongly bound with duplicated sections. Returning it to the bookshop he meets a girl with a similarly defective copy, and together they set out to finish the story, which becomes ever more complicated. A bewildering sequence of partial novels by different (or not) authors about sex, war, spies, all totally gripping, none ever proceeding to an ending. It's not another book about the dreaded abstraction 'narrative', but about the material existence of books, which had been on my mind for two reasons.

Back in Idealog 227, 2013, I recounted my adventures trying to publish my book Sampling Reality, about the intersection of philosophy, information theory and neuroscience – as a Kindle Book, using a then state-of-the-art conversion tool called Calibre. I was beset by problems like all special characters displaying as '⎽' and suffered agonies trying to get table-of-contents to work, employing every conceivable different input format like docx, pdf, rtf, odf and more. I eventually got it accepted, and it’s since failed to disturb the best-sellers list, but at least it’s there.

Reflecting on this a few weeks ago I’d decided perhaps to give the book away as a PDF, when almost immediately after this thought occurred to me, Kindle Direct Publishing emailed to announce that they’ve just launched a paperback book print service in addition to ebooks. Too good to ignore, I set about downloading the new version of their production tool Kindle Create. It’s available for PC and Mac only so I installed it onto my Windows laptop, which has lain unused except as a print server for several years, and has a developed a sporadic problem involving a locust-like system process called ‘Runtime Broker’ that gobbles CPU and memory until I squash it in Task Manager.

I’d far rather have done the conversion on my Chromebook, but I soldiered on in semi-crippled Windows and it went better than Calibre, give or take fixing the odd missing contents item. I uploaded the resulting .kcb file to KDP, only to be curtly told it wouldn’t fit into the default 6” x 9“ page size. I’d never considered page size with my ebook of course. Having seen enough of Task Manager, I went downstairs to my Chromebook and reformatted the original .docx in Google Docs, emailed it to myself and went back upstairs to re-do the Kindle Create, which was now accepted. Made a new cover using Kindle Cover Creator, which went well, submitted it all and 72 hours later was given a PDF proof and told my paperback, now 191 pages long rather than 155, has been published. Result.

The second bookish adventure involved reviewing Thomas Piketty’s 1093-page Capital And Ideology for The Political Quarterly. The review copy arrived and is the size of two house-bricks and at 1.7kg almost as heavy. My normal mode of reading, flat on my back on the sofa, was entirely out of the question, and even propping it up on the table was problematic - laid flat it won’t stay open at your page, propped up it slides downwards. I purchased a neat little lectern on Amazon, made in Germany from bamboo and chromed wire.

The book’s spine width and weight were too much for it until I bent the wires into a different shape, and then constructed an ingenious – if I say so myself – system of rubber bands and grommets to hold the pages open. After a week or so of this I contacted Belknap and asked whether they had a Kindle edition. They didn’t but their charming UK PR sent me a PDF instead. It’s also big, at 25Mb, and unusably slow in Adobe Reader, but fortunately I have a better PDF viewer, the marvellous Chrome extension PDF.js. A community-driven GitHub project, built with HTML5 and available free from the Web Store, it’s maintained by Rob Wu (PDF Viewer Chrome extension) who describes it thus: “Our goal is to create a general-purpose, web standards-based platform for parsing and rendering PDFs.”

And oh boy does it parse, searching the huge book fast enough to be my principal means of navigation, and its page zoom is far nicer than Adobe’s too. Over the months my review took I barely touched the printed book again. So what exactly is the postmodern moral of this story? Are e-books finished, are they still the future, or are they just yet another tool that has its place when applied to the right problems?

JUST A PLACEBO?

Dick Pountain /Idealog 314/ 06 Sep 2020 10:37


It feels as though the Covid pandemic has made virologists and epidemiologists of all us (except those who deny it’s happening of course). Suddenly people are following vaccine development with sport-like attention and discussing the difference between cytokine and bradykinin storms at the bus-stop. Or perhaps that’s just my circle of hypervigilant friends…

This interest is of course highly practical, since returning to any semblance of normal life depends to a great extent on success in a vaccination program. But I sense there’s another dimension to it: it makes us feel better when we understand what’s happening to us, gives us a sense (perhaps illusory) of control. This effect of increased confidence can in some cases affect our real bodily functions, when it’s referred to as ‘the placebo effect’.

Medical science has only fairly recently started to take the placebo effect seriously, and its power appears more remarkable the more is discovered. The effect was dismissed for a long time for the very good reason that it appears to conflict with the central dogma that separates science from magic, namely that the mind cannot *directly* affect matter (without which we’d still be using Eye Of Newt instead of dexamethasone). That’s changing as we learn about the real material pathways that exist between software processes in the brain (that is, thoughts) and bodily processes. These pathways are mostly chemical rather than electrical, depending upon hormones and neurotransmitters distributed via the bloodstream. Incidentally, this is one more reason why the pursuit of Artificial Intelligence will remain stunted so long as it treats intelligence solely as a computational function of the brain, ignoring the intimate two-way communication between brain and the rest of the body’s organs.

Acceptance of the placebo effect grew with the pharmaceutical industry as it introduced drug trials and discovered that a placebo (from the Latin ‘to please’) – that is a fake pill, often just sugar – could sometimes produce an effect similar to the real drug. At first explanations were purely psychological, concerned with expectation: if you *expect* a pill to cure your headache then it might. The placebo effect was and remains a problem for drug trials, since untangling it from the real drug effect is difficult. There’s also an opposite, ‘nocebo effect’ where patients who are informed of possible side effects of a drug can experience or intensify them.

Psychological explanations raise the difficult question over whether such effects are ‘merely’ imaginary, or are physical, a question that also arises about illnesses that medicine suspects may be ‘psychosomatic’. The emphasis nowadays is shifting, not to discount psychological explanation entirely but to reveal how mind actually affects bodily systems. Placebo analgesia occurs when the mental expectation of relief stimulates the limbic system to release hormones called endorphins which behave like opiates. Expecting an antidepressant effect can release dopamine and thus improve mood, ditto for insulin and blood sugar or vasopressin and blood pressure. It’s even been demonstrated that patients can be conditioned, Pavlov-style: administer a drug in a drink with a distinctive taste, and if the drug is removed merely tasting the drink may produce its effect. In short, the placebo effect reveals yet another bodily system – like the immune, muscular and gastro-intestinal systems – that functions as a computational control system separate from and in parallel with the brain.

These placebo pathways can be trained and nurtured to an extraordinary extent, so that they become the basis of practices like yoga and acupuncture, or even the source of what would have once been called ‘miracles’. Wim Hof, a 60-year old Dutchman, can immerse himself in ice water for 45 minutes and swim under 50 yards under the ice of a frozen lake, by conditioning his breath control. Free divers train themselves into feats of breath-holding approaching 20 minutes. Similar feats of ‘mental’ pain relief are routinely reported both in wartime and among marathon runners. An Italian placebo scientist called Fabricio Benedetti gave weightlifters what he told them was a performance-enhancing drug, actually a placebo. He also secretly gave them lighter weights, which convinced them that the drugs were working. When he surreptitiously replaced the normal weights, the muscular force they were able to exert increased while their perceived fatigue remained the same.

I suspect there’s a lot more yet to discover about the placebo effect and its pathways, with enormous consequences not just for medicine but for sport, and everyday life. There’s already evidence that it can even stimulate the immune system, though whether it could ever be trained to resist infections like coronavirus seems pretty unlikely at the moment. Such slim hopes are certainly no excuse either to slacken the effort to develop vaccines, nor to stop persuading anti-vaxxers that they will need to accept them.

SHOW MUST GO ON

 

Dick Pountain /Idealog 313/ 17 Aug 2020 10:32

Last night I ‘attended’ a superb jazz session by world-class musicians Wayne Shorter, Herbie Hancock, Dave Holland and Brian Blade. Admittedly it happened 16 years ago – in the Große Konzertscheue, Salzau, Germany -- but I heard it almost as well, and saw it far better, than if I’d been there. My sofa is easier on the bum than most hall seats. It was a free YouTube video, in HD quality, which I Chromecast to my LG smart TV that pushed the sound through my vintage hifi system (excellent though hardly audiophile: Denon amp, Castle speakers). The camerawork was exceptional, in the German manner, so I saw the players’ fingers on their instruments and facial expressions in a close-up never experienced at a live gig. I also avoided queuing for the cloak-room, and being surrounded by people eating steak and chips and chattering instead of listening. So was this virtual concert a satisfactory replacement for the real thing?

I won’t go all gushy about the excitement of travelling in anticipation, about sharing an enthusiasm with other warm, breathing human-beings (which was once true) but will instead focus on more pragmatic considerations. First off, those musicians got as good as they are through a lifetime of playing to live audiences in clubs all over America and Britain (Holland is English), being paid a pittance by tight-fisted promoters. Are kids coming up today via Logic-Pro-on-bedroom-laptop and social media going to develop similar or equivalent skills: only time will tell, but many YouTube channels suggest perhaps not.

Secondly, can a viable music scene be maintained through payment for online performance? I didn’t pay for that Salzau video, and had YouTube been charging I probably wouldn’t have watched it, not knowing how good it was going to be. On the other hand I frequently pay £40+ a ticket to see acts at the South Bank, Jazz Cafe or Ronnie Scott’s. I don’t know what percentage of that gets to the musicians, but nor do I know what slice (if any) of YouTube’s ad revenue went to them for that video.

This applies even more so in the world of classical music. During lockdown in June I watched a week of excellent lunchtime concerts streamed from the Wigmore Hall, including a staggeringly fine ‘Winterreise’ by Mark Padmore and Mitsuko Uchida. The visibly empty seats brought home frighteningly just what the virus is doing to us. As regular attendees at the Wigmore we like to sit stage-side for which we pay £12 to £20 a head. I didn’t pay that for all those streamed concerts, though I did make a one-off donation.
The brutal truth is that the psychology of paying for streamed entertainment is very different from paying for live entertainment. Rightly or wrongly, you are unlikely to pay as much to watch from your own sofa, providing your own refreshment, as you would to travel to a special event at a concert hall or club. And even the alternative ways to pay for online entertainment can be fraught because of the distinction between pay-per-view and subscription.
Streaming has two huge advantages, instant access without travelling, and a vast repository of past performances. Instant access can make it possible to sample performances that you wouldn’t normally consider, and hence be educated and change your tastes - but only if it’s cheap enough that quitting ones you dislike after a minute or two doesn’t hurt too much. That  

was the difference between Spotify and Apple’s now defunct iTunes. I’m happy to pay £10/month for Spotify Premium which I use every day, not just listening to favourites while walking, but for finding new music or researching all versions of some tune. I wouldn’t do any of that were I paying per track.

Movies and TV are more complicated. I don’t subscribe to Netflix, Prime, Hulu or the like, because they don’t have enough of what I like to justify another monthly bill. But I do buy or rent one-off showings of movies – for example hard-to-find oldies like ‘Tampopo’ or ‘Babette’s Feast’ – and I use BBC iPlayer and All4 to binge watch series (I’ve pigged out on all seasons of ‘Line Of Duty’, and all seven years of ‘30 Rock’).

If coronavirus changes the way we consume entertainment forever, market forces alone are unlikely to save ‘the talent’. The print publishing industry faced this problem for years over library lending, and they came up with PLR (Public Lending Rights) and ALCS (Authors' Licensing and Collecting Society) which collect royalties on behalf of authors fairly efficiently. I suspect similar institutions will need to be cobbled together to collect revenues from online service providers on behalf of musicians, and even perhaps starving Hollywood moguls (joke alert)...

END OF MOORE’S LAW (AGAIN)

 Dick Pountain/Idealog 312/15 Jul 2020 01:58

It’s now a good 11 years since I last wrote a column (Idealog 180) about the end of Moore’s Law, in which time the number of transistors on a chip must have grown at least 100-fold. As I said in that column (just as in the one four years earlier), predicting the end of Moore’s Law is a mug’s game. It’s a bit like predicting The Second Coming or the arrival of a Covid-19 vaccine. In a 1997 article for Byte I’d predicted that lithographic limits and quantum effects would flatten the curve below 100 nanometre feature size, and I was only off by one order of magnitude which almost counts as a win in this futile race. Intel’s latest fabrication plant, built to produce chips with minimum 10 nanometer feature size was very late indeed and only started delivering chips in 2019, five years after the previous 14nm generation of chips.

However in the last few months a chorus of highly-qualified commentators have been declaring that this time it’s for real: high performance computing pioneer Charles Leiserson of MIT has remarked that “Moore’s Law was always about the rate of progress, and we’re no longer on that rate.” It’s not just those physical limits on feature size I was writing about, but economics too. The cost of building a new fab has been rising 13% year-on-year and is headed north of $16 billion, at precisely the time that Donald Trump, great tech entrepreneur that he is, is calling for US companies to bring chip fabrication back home as part of his trade-war on the Far East. Only Intel, AMD and Nvidia can even contemplate a next lower level of feature size (and Nvidia’s not all that sure).

Of course reaching bottom in feature size doesn’t mean the end of all progress in computing power. The effect of Moore’s Law itself encourages software bloat - why bother writing efficient code if next year’s chip will speed up today’s crappy code. This is a problem waiting to be tackled: most of today’s commercial software could probably be sped up 100-fold by a decent rewrite. But another problem is that rewriting code is almost as expensive as fab-building.

Parallelism looked like the solution for a long time, and it sort of was: even the cheapest mobile phones today use multi-core processors, and AMD is selling 16-core chips right now. The thing is, the more cores you add to a chip, the more of the silicon real estate gets eaten up by interconnect, and what’s worse, the model of parallelism employed for x86 family processors isn’t automatically exploitable by old software without a rewrite.

There are two hugely different groups of people who need the extra power of multiple cores: games vendors and AI developers. The former have the cash to rewrite their flagship games for each CPU generation. The latter need far more, and far more flexible, parallelism than these chips offer, and so are headed off along a path toward special-purpose processors (I wrote about some of their novel architectures in Idealog 301 after the 2019 CogX show). Such ‘IPUs’ can speed up the kind of massive matrix and convolution calculations performed during deep-learning by several orders of magnitude - the problem is they can’t run Animal Crossing, or Google Chrome, or Microsoft Word. They are not general-purpose processors. They are however potentially incredibly lucrative, since they will eventually end up in every mobile phone, Alexa-style interface or self-driving vehicle. Venture capital is queuing up to invest in them just as mainstream processors begin to look less like a hot tip. Neil Thompson, an economist at MIT’s AI centre, has just written a paper called “The Decline of Computers as a General Purpose Technology” which gives you an idea of the drift.

Moore’s Law is underpinned by the scaling behaviour of CMOS fabrication technology, and it’s that we’re approaching the end of. Professor Erica Fuchs, of the engineering and public policy at Carnegie Mellon university, worries that a successor technology with equally benign scaling properties, that could maintain Moore’s Law for general-purpose chips, is as yet unknown and may take years of basic research and development to find with no guarantee of success. Candidates might include carbon nanotubes, graphene transistors, spintronics or even the dreaded qubits but none of these are obvious replacements for CMOS scaling. She calls for a huge boost in public research funding to replace all the venture capital that’s being diverted into special-purpose AI chips. Unfortunately the colossal cost of the Covid-19 pandemic is likely to make that a very hard sell indeed, given that most politicians have little idea of what chips do at all, let alone the subtle distinctions between special and general purpose ones.

[Dick Pountain plans to leave it for 16 years before he writes another Moore’s Law column]

BAND OF BIG BROTHERS

 Dick Pountain/ Idealog311/ 9th June 2020 16:14:55

Have you spent time during lockdown worrying about what life will be like after this pandemic ends? For me personally lockdown itself hasn’t been a huge deal as I’ve been working from home for best part of 30 years, but I’m noticing other changes already. One example, this time last year I devoted two of these columns to an AI conference, CogX, which was held a short stroll down the canal from my home in King’s Cross. Well, I’m there again this year as I write this, but virtually via live streaming. That works well enough, in fact it’s easier to hear the speakers and read their slides than it was sitting in a tent in the rain. But I can’t help notice that there aren’t so many big hitters from US labs on the roster this time. Surely that couldn’t be because an invitation to sit on your sofa with a laptop isn’t so attractive as a free air trip across the Atlantic?

Actually the loss of air travel worries me less than most too, because 12 years of too-and-fro from Italy already squelched much of the romance of flying for me. What worries me more is our increasing reliance on online purchasing, while the only shops open were food shops. And it’s not just the way Amazon is taking over ever larger chunks of the retail sector, it’s that all of the US digital giant corporations are garnering profit and power from the global lockdown.

I’ve never been one to promote conspiracy theories myself, and as I said in last month’s column I’ve willingly placed myself under the tutelage of Google to a degree that would make some people nervous. But there’s no escaping the fact that Amazon, Google, Microsoft, Facebook, Uber and the like all have ambitions to provide services which in many parts of the world are either monopolised or heavily regulated by the nation-state. These corporations are all working hard to penetrate the healthcare, education, security and transport sectors, by providing innovative and ‘disruptive’ services that parallel what they’ve achieved in the retail, entertainment and social media sectors – and there’s no doubting that the extraordinary infrastructures and AI capabilities they’ve all invested in are indeed more efficient than many, even most state equivalents. I’ve often wondered how the NHS would look if it had a platform of the efficiency of Amazon’s to connect patients, GPs and hospitals.

Trouble is, they remain unelected, commercial enterprises who have no other commitment than to their shareholders, and it’s unthinkable – even to extreme free-market libertarians like our current government – to give them such massive control over our economy. Another problem is that they notoriously evade paying fair taxes in the territories they operate, thus depriving the public coffers of funds needed to compete with their services. There’s a potential solution to that, short of nationalising them or taxing them into retreat: governments with sufficient resolve could strike deals where these corporations pay a fairer tax whack in part, through partnerships that offer their platforms free to improve existing public services.

That however would require the state itself not to be evil, and seeing how some are already applying digital tech to integrate welfare, security and taxation systems isn’t encouraging. India’s Aadhaar system, Singapore’s ‘Smart Nation’ and Alibaba’s cooperation with Chinese local government to run their Social Credit system all present potential threats to personal

liberty. There are questions over how well these systems actually work, but they certainly grant the state a scary degree of extra knowledge and power over citizens: participating in a demonstration, perhaps even voting for an opposition party, can be punished by loss of benefits.

While George Orwell was writing ‘1984’ the first modern 625-line television system was being introduced (in the Soviet Union as it happens) and he foresaw the effect this new electronic medium would have on authoritarian societies. However because TV is only a one-way channel, his picture of Big Brother’s regime was somewhat stunted. Perhaps had Orwell lived to see our two-way, social media, internet, he might have concluded it had democratic potential? Perhaps he would, but we know better...

Deepfakes and disinformation are neutralising whatever democratic potential the net might have had (pioneered as it happens by those same Russians: Putin’s KGB background gives him a tech savvy that’s notably lacking among Western leaders). The virus keeps us all at home, staring into our screens, shopping, and wallowing in streams of false information. In 1988, Guy Debord nailed it in ‘Comments on the Society of the Spectacle’:

“Networks of promotion/control slide imperceptibly into networks of surveillance/disinformation. Formerly one only conspired against an established order. Today, conspiring in its favor is a new and flourishing profession.”

[Dick Pountain oft times quests through the deep, dark labyrinth of Amazon’s menus to slay a free Prime membership on its last day]