Doctor Strangememe Or: How Oliver Learned To Stop Worrying And Love AI

We see lots of “You can use AI for that” boosterism coming from the corporate right, desperate to convince investors in this stuff that the money they’ve thrown at generative AI will pay huge dividends. So I was bemused to see one Oliver Markus Malloy, apparently a progressive Democrat bigging it up (link to his piece here).

His core argument is that reactionaries (MAGA, its global network of far right allies and weird techbros) use this stuff, so progressives also need to use this weapon or get destroyed when the right brings its AI gun to a political knife fight. It’s classic arms race thinking - just as the rival imperial powers threw steel and sweat into the pre-World War One dreadnought-building race, he argues that progressives should respond to their rivals' AI by throwing scraped content into producing bigger, badder memes than the enemy (he also makes another argument that content created by AI can be art, not just derivative "slop" but it's his thoughts on political comms that concern me most).

I’m sceptical about the “this is an arms race we must win” argument. OK, we’re agreed that the corporate right/far right alliance has an advantage when it comes flooding the zone with clickbaity content quickly and at scale. The argument goes that by not adopting the same techniques & tools, progressives are ceding the information space to their enemies.

I'm not convinced that progressives should mirror their opponents in this reactive way. I'm not arguing that the far right doesn't use this stuff, or that the sheer volume of content it allows them to pump out doesn't work. After all, "Quantity has a quality all its own" as Josef Vissarionovich probably never said.

But before we decide that "we need some of that", maybe we need to think about what lessons we can learn from the last time the right used IT and big data to make historic gains. In the run-up to the United Kingdom's vote on whether or not to leave the European Union and Trump's first battle for the White House, we saw how bad actors like Cambridge Analytica were able to use big data to flood social media with vast quantities of targeted disinfomation and rage bait. 

The likes of Facebook and Twitter (as it then was) were overrun with trolls, bots and a blizzard of vaguely plausible but disingenous messages. Many of the messages contradicted one another but, because they were targeted at specific groups, members of other demographics who saw different messages didn't see the contraditions.

This IT-enabled flim-flam certainly played a role in selling different groups of the unsupecting voters a bill of goods (the mis-sold product being Brexit in the UK and Trump in the US).

So, what would've happened if progressives had a Cambridge Analytica equivalent of their own in the run up to the 2016 EU referendum and US presidential votes? Could they have swung the results in different directions? I'm not so sure, because what we were looking at then wasn't just a gap in capabilities but a gap in content and ethics. 

Remember what I said about micro-targeted and often contradictory messages? Yes, delivering those messages was a tech problem, but the content of the messages was as important as the medium. Tech aside, the innovation was to efficiently send contradictory messages tailored to appeal to different sets of voters and to amplify existing misinformation which was already out there in the wild, courtesy of years of mendacious campaigning in the right wing press. Think about that for a moment.

Using that tech wouldn't have worked (at least in the same way) if the people using it had been held back by the moral scruples to make an honest case. To use the example I'm most familiar with, the EU Referendum and the effectiveness of the Leave campaigns:

...But, as I discovered while knocking on doors during the campaign, many Britons believe all sort of bizarre things about the EU that have no basis in fact, and the source of which is ultimately newspapers – for example, that most immigrants to Britain come from the EU, that 20 per cent of the population are EU migrants or that 75 per cent of Britain’s laws are made in Brussels.

Before and during the campaign, the eurosceptic newspapers carried a strong message that EU migrants were causing enormous problems in Britain. They ran front-page after front-page of scare stories about how migrants and refugees were trying to get into the country – often conflating the two groups. Many of these articles were factually incorrect. Even on the day after the Orlando shootings in Florida, the Daily Mail – uniquely among UK papers – led its front page with ‘Fury over plot to let 1.5 million Turks into Britain’. The written press did a great job in reinforcing Vote Leave’s twisted message that thousands of foreigners – whether asylum-seekers, Romanians, Syrians, terrorists or Turks – were all hell-bent on entering the country.

From a piece by Charles Grant of the Centre for European Reform.

Ultimately, the Leave campaigns' ability to leverage their big data advantge rested to their willingness to lie shamelessly without fear of retribution. As Charles Grant put it "They exploited the fact that in political advertising, unlike commercial advertising, there are no penalties for untruths".

So I'm not sure that, even with the same resouces, the other side would've been able to win big by upping its big data/social media game. Unless, of course, it was willing to lie as shamelessly itself. The tech was just the delivery system. The weapon itself was made of good old-fahioned lies and bullshit and its effectiveness relied on our old friend the bullshit asymmetry principle, AKA Brandolini's Law ("The amount of energy needed to refute bullshit is an order of magnitude bigger than that needed to produce it").

There are two main reasons why I'm not pro-using the latest tech tools to "flood the zone with (our) shit" in an effort to emulate how the bad guys have been winning so far. The first reason is that it's lying and that's just straight wrong.* Secondly, it moves the fight onto the enemy's territory. If the opposing party has no qualms about lying on an industrial scale and you decide to get into a lying contest with them, you're going to lose. 

You'll lose because blurring the lines between lies and truth is more helpful to liars than it is to members of the evidence-based community. So long as there is still some shred of evidence or objective truth out there by which people can judge them, liars and propagandists are vulnerable. When both sides are unreliable and there's no objective standard of truth, you're into a world of "he said she said" and "they're all as bad as each other".** And, aside from direct disinformation, getting there is a win for bad actors from Cambridge Analytica to MAGA influencers to Russian troll farms. Yes, we're circling back to that famous Garry Kasparov quote:

The point of modern propaganda isn't only to misinform or push an agenda. It is to exhaust your critical thinking, to annihilate truth.

In other words, creating a space where there are no objective standards to which the powerful can be held or which they can judged by. Believe me, you don't want to go there.

So how does this map on to the latest wave of AI-driven political comms? Well, the next image seems to have been a viral anti-MAGA hit. It allgedly upset the MAGA regime so much it led to a Norwegian tourist being refused entry to the Land Of The Free™ for having it on his phone (I think the US side eventually denied that this image was the reason but, honestly, I can wholly believe tat the MAGA crowd are thin-skinned and control-freakish enough to have barred someone for making fun of them):

Meme of J D Vance as a bald, bearded baby

 

And you'd need a heart of stone not to appreciate those pompous assholes being mocked, or the Streisand effect that kicks in when they try to suppress the mockery. 

The problem is ... as far as I know, the viral J D Vance memes were produced using Photoshop, not AI.

As far as I know. But even if they were the products of AI, would reaching for AI be worth it? Clearly you could do this in Photoshop or similar without colluding with the damage AI is doing to the environment, wholesale copyright theft, disrupting creative jobs (no, techbros "disrupting" isn't automatically a good thing - while it may be good to disrupt, say, a terrorist cell, disrupting the lives of people who are just trying to make a living without harming anybody else is a bad thing), widening the already obscene inequlities of wealth and power still further, all in the service of what may well be yet another a very expensive financial bubble.

And that's without considering that AI is asymetrically attractive to bad actors who want to generate big volumes of content quickly without being concerned by troublesome details like truth or ethics. Deepfakes, amplifying existing prejudices (even when there isn't a man like Elon "Roman salute my arse" Musk behind the curtain obviously tweaking his pet AI to let it express its inner MechaHitler)...

... the list of opportunities for plausibly deniable deception & generally messing with people's heads is long and depressing. What I'm seeing is another tech-enabled opportunity for people with no principles to flood the zone with shit in much the same way they did a decade ago. The opportunities for progressives and people who value strightforward messages anchored in reality and something approaching objective truth seem a lot more constrained.

 OK, beyond dumping on AI do I have any thoughts about what would work? Well, the ten years since the last tech attack seem to have been largely years of lessons not learned. But not everywhere. In Finland, they got the memo about going for the harmful content, actively pushing back against fake information, fake images and clickbait being used to attack the principle of informed consent with a programme of education and public infomation aimed at children and adults alike. Another lesson not learned in too many places by too many politicians is there are plenty of progressive policies which are poplar and can cut through, even against the screeching coordinated right wing media claque, if they're presented clearly and with conviction. Just ask Mr Mamdani. And remember the lesson of the J D Vance memes; these militant far right characters are freakishly weird. Stop normalising their batshittery and mock them - it's easy when you try and the best part is that we're quite smart enough to do it right now with without the help of AI.

 

*Yes, I know that technically there are cases where it may be right to lie (e.g. lying to a knife-wielding maniac about the whereabouts of his intended victim) but that's not the sort of scenario we're considering here.

**This both-sides-ism already happened during the Brexit debate - even though the vast majority of points made by the Remain campaign were broadly correct and borne out by subsequent events, the reaction of the Leaver campaigners to their lies being called out was to find a dodgy statement from one of the millions of Remainers (or to make one up) and shout "See! Both sides!" blithely ignoring the fact that they'd been caught red-handed themselves.

Ecuador the recursive

Notice anything odd about the flag of Ecuador?

Apart from the busyness of the design, that is, and the now dodgy associations of the fasces. To be fair, there are a lot of cluttered flag designs featuring a coat of arms, and Ecuador's isn't the only coat of arms to still feature a fasces - there's one on Watford Borough Council's coat of arms, too. 

Watford Borough Coat of arms (image credit Heraldry Wiki contributor Knorrepoes)

 

Spooky coincidence; the fasces-bearing arms of Watford Borough Council were granted in October 1922, the very same year and month when Mussolini marched on Rome and then took power in Italy.

Anyway, back to Ecuador's flag, and the oddest thing about it (in my opinion). Look at the coat of arms in the middle.

The odd thing isn't the condor on top, the fasces on the bottom or the shield (?) depicting a sun, mountain, river and ship in the middle. It's what flanks the design; halberds from which are draped ... Ecuadorian flags. It's a flag containing little copies of itself. A recursive flag. A bit like the Droste effect, only not quite, with more than one small copy.

I'm now idly wondering whether there are any other flags which contain images of themselves. I think there probably must be some (although not necessarily national flags). My very quick web search hasn't turned any up, although it did come up with some recursive flags identified by Redditor and vexillology enthusiast Cawren. The trouble is, Cawren's definition of recursive wasn't quite what I was looking for; the flags cited are ones where you can zoom in infinitely, keeping the same aspect ratio and the flag will stay the same, i.e. a very boring fractal, as illustrated below.

 

But it's flags which literally contain pictures of themselves that I'd like to see and some time when I have more time and energy I'll devote more than a couple of minutes to looking for more examples. Or maybe I won't. Who knows?

 

Watch this space. Or don't.

 

Anyhow for no other reason than it seems appropriate, here's my favourite remix of that 1997 banger Ecuador, originally a hit for the DJ and record production team Sash!* (it got to number one in Flanders, Romania and Scotland). 

 

Enjoy.

.

*I should point out, for those too young or old to be aware, that the exclamation mark is part of Sash!'s name, like Westward Ho! But the asterisk isn't. Obviously.

Secrets and spies

So the Pentagon leaker, turns out to have been (apparently) an immature nerd trying just to impress a bunch of teenage boys with his edgelord antics,* rather than a guy who sold secrets for money, or a Philby/Maclean-style sleeper agent recruited by ideological fellow travellers from abroad.

The aspect I find more interesting than the confused motivations of a messed-up 21 year old was raised in this piece by Daniel Drezner. Drezner lists the material that we know has been leaked and concludes that it seems pretty "meh":

You know what? I’m not seeing much in these reports that I find particularly surprising or shocking. All of these assessments mirror the takes one would get on each of these questions from analysts with zero access to classified intelligence.

In other words, the Kremlin could have deduced most of what Teixeira leaked from open source intelligence (OSINT), so having a mole in the other side's camp doesn't necessarily give an opponent that big an edge.

Coincidentally, John Quiggin has reacted to the leaks by reposting a piece he wrote in 2003, in which argues that the effectiveness of spies and spying in general is very much over-rated.

 

The basic lesson of game theory for a game of bluff like that of espionage is that, as long as it is possible for counterspies to generate misleading information most of the time, spies are useless even when their information happens to be correct. If the defence plays optimally, the spymaster can never have any reason to believe one piece of information produced by spies and disbelieve another.

The biggest problem isn't, though, that spying is mostly ineffective, but that it provides the justification for having an intrusive intelligence apparatus which ends up being deployed against the state's own citizens rather than against hostile foreign powers:

The spy myth clearly served the interests of intelligence agencies, which prospered during the 20th century more than any set of spies before them. The real beneficiaries, however, were the counterintelligence agencies or, to dispense with euphemisms, the secret police, of both Western and Communist countries. The powers granted to them for their struggle against armies of spies were used primarily against domestic dissidents.

Quiggin's critique of intelligence services over reach seems as relevant to me now as it was when he wrote it in 2003, in the wake of 9/11. Perhaps it's even more relevant now when the organs of the state are partnering up with hugely well-resourced surveillance capitalists to keep an ever closer watch on their own citizens.

*If Teixeira had been a 51 year old billionaire, rather than a 21 year old national guard, he could have just bought Twitter to impress his teenage fanboys.

 

 

Halloween photoblogging: post-apocalyptic holiday resort

Here's some spooky photoblogging just in time for Halloween. I'm just back from a week's break in Croatia, at a small resort near Dubrovnik. One bay up from where we were staying was another bay with a wide curve of beach and a massive hotel complex ... all of it abandoned and crumbling, with some of the facades pockmarked by gunfire. It was properly creepy after the zombie apocalypse stuff.

If this has been the in the UK the whole site would have been closed off with stern warning signs, fencing and barriers to keep the curious out, but this lot was just left open for the curious to wander round and explore. I figured it was atmospheric enough to be worth a few photographs, so here they are. On first visit I had no idea what had gone on here, or why the whole complex had been abandoned. A quick internet search uncovered the back story, as told by the good folk at Atlas Obscura - click the link here for details of the abandoned hotels of Kupari and how they came to be abandoned.

 

Baiting your hook with virtue

I've been fascinated, in a slightly horrified way, by how right wing propagandists constantly try to reframe public discourse by the repetition of a small vocabulary of loaded words and phrases. The pejorative use of "woke" is seems to be the most common of these linguitic tics right now, but accusing your opponent of "virtue signalling" is still widely used as a catch-all way of dismissing an opposing view.

These sort of prefabricared buzz phrases fit together like Lego blocks, so you don't have to go far on social media before you trip over people complaining angrily about "woke virtue signallers". The point seems to be to keep using the same linguistic cues, over, and over, and over again in that hope that, thanks to availabilty bias, these negative connotations will stick in people's minds like mud.

The assumptions underlying this sort of language are, of course, highly questionable. What does "woke" really mean, and is it necessarily a bad thing? Are all examples of alleged "virture signalling" really always examples of people insincerely paying lip service to virtuous behaviour merely to look good? These are the things the propagandists using this sort of language doesn't want you to think about.

None of this stuff is new, but I did come across a new phrase that underlines how incoherent the right's weaponisation of "woke" and "virtue signalling" is. I was listening to a recent episode of Michael Rosen's Word of Mouth* radio programme about language and how we use it. The episode dealt with the language of online dating and introduced me to a new phrase I'd never heard before; "wokefishing".

 Wokefishing is a more specific form of "catfishing", which is deceptively creating a fake online identity to lure somebody into a relationship under flase pretenses. In wokefishing, the romance scammer pretends to subscribe to a set of what people are shorthanding as "woke" beliefs, which could cover things like social justice, women's rights, minority rights, tolerance, green issues and so on, in order to lure somebody into a relationship.

Putting this new word next to the way the right weaponises "woke" and "social justice" prompts some interesting thoughts:

Firstly, culture warriors on the right want us to see "woke" as a pejorative term of abuse. If people are pretending to be woke and using woke identities it to make themselves more attractive to potential partners, this suggests that being woke is an attractive trait in a dating profile, a plus, like interesting, good sense of humour, honest, etc. So not only is it unclear how being actively opposed to social justice is a bad thing, as the "anti-woke" brigade claim, but wokeness is also a plus on a more superficial, social capital, level.

Which brings us, secondly, to the crime of "virtue signalling". In a literal sense, that's what wokefishers are doing; pretending to believe in virtuous things for selfish and cynical ends, while actually not believing in them at all. So who is doing the "virtue signalling" here? By definition, the people who are wokefishing aren't woke. Woke people saying that they subscribe to woke beliefs, or even better, honestly saying that they live by those beliefs and try to advance their principles in the world aren't "virtue signalling", they're just being honest about their core beliefs.

So we've got that rare beast, a genuine example of virtue signalling in the wild. And the culprits are, by definition, people who aren't woke. 

I would say that this is an example psychological projection from the right's keyboard warriors. It sounds good, but I'd be falling into the same trap as the people who routinely use "virtue signalling" as a taunt; using a superficially sciency, technical-sounding term that's really quite sloppy and imprecise. After all, the term psychological projection comes out of Freudian analysis, a field which doesn't exactly come top of the class when it comes to having a rigorous evidence base. 

 Not psychological projection, then. An older name will do. Hypocrisy.

 *Link to the episode on BBC Sounds here. Also available on at least one podcast app (I listen to Word of Mouth on Stitcher)

 

"Anti-wokism" - back to the bad old days

According to Elon Musk and Jordan Peterson "wokeism" is a "mind virus". Before stepping down as Conservative party chairman, Oliver "the privet hedges of a free people" Dowden dunked on "wokeism" at greater length (but with no greater clarity) in a speech to the hard right US Heritage Foundation:

“Rogue states are seeking to challenge the international order. And at the precise point when our resolve ought to be strongest, a pernicious new ideology is sweeping our societies...

...It goes by many names. In Britain, its adherents sometimes describe themselves as ‘social justice warriors’. They claim to be ‘woke’, awakened to the so-called truths of our societies. But wherever they are found, they pursue a common policy inimical to freedom.”

The use of "woke" as a catch-all snarl word in the right's culture wars is relatively new. The substance of the fight they're picking isn't. I was reminded of this when I came across an old Daily Express cartoon that somebody had posted on Twitter. A picture is worth a thousand words, but I'll repeat what I tweeted just to hammer home what's wrong with this picture: 

The Conservatives picking a fight over Culture War talking points is nothing new. See this Express cartoon from the 80s which did the same thing back when people could openly attack things like anti-racism without even hiding behind the euphemism of being "anti-woke".

I think a lot about this cartoon. Especially the upside-down figure labelled "council power". What does that even mean? Vampiric local authorities sucking honest Tory ratepayers dry? And hammers & sickles to label everything they don't like Marxist. Batshit then & now.

There's not much to add to this - the war against "wokeism" is a war against social justice, solidarity, equality and inclusiveness. It's a war fought for the attitudes encapsulated in this cartoon, a war for division, for othering, for punching down and keeping the designated outgroups firmly in their place. 

The war starts with words and memes, with mockery and humiliation. The end point is action. And, if you still haven't joined the dots between images, words and action, remember the Heritage Foundation, the "think tank" (lobby group) that Oliver Dowden crafted his "anti-woke" talking points for? 

 

Since its founding, the Heritage Foundation has become “inextricably intertwined” with Republican administrations and lawmakers in Washington. Heritage brands itself as a beacon of the intellectual conservative establishment; in reality, it is an organization that regularly spouts hateful ideas on par with organizations like the Family Research Council (FRC), which has earned designation from the Southern Poverty Law Center as a hate group. Much like its peer groups, Heritage dedicates significant energy to extremist policy recommendations that hinder access to abortion and birth control and promote discrimination against LGBTQ individuals...

...Kay Coles James became president of The Heritage Foundation in the beginning of 2018; she had served as a Heritage board member for over a decade. James previously worked in the George W. Bush administration and as Virginia’s secretary of health and human services. James has compared LGBTQ people to “drug addicts, alcoholics, adulterers, or ‘anything else sinful,’” and has also tweeted that “abortion is a form of discrimination,” attempting to paint her anti-abortion work in the same vein as fighting racism.

Pro-Lies.org
 

It starts with words. It ends with action. It ends with reversing decades of hard-won progress overnight. Goodbye Roe vs Wade, it was nice knowing you...

Cartoon apes want to be free.

Bitcoin, ethereum and other major cryptocurrencies have been hit by a huge crash over the last week, partly triggered by the shock collapse of a major coin.

The bitcoin price has lost 25% over the last month with its biggest rival ethereum down over 30%.

Other smaller cryptocurrencies have been even harder hit—sparking fears others could collapse entirely.
 

Now, as serious economic "shock therapy" warning signs flash, analysts at Wall Street giant Morgan Stanley have predicted prices of digital collectible non-fungible tokens (NFTs) could come under pressure. (Forbes)

As a complete outsider, one thing occurred to me, once I'd picked myself off the floor from laughing myself stupid at the plight of people whose idea of fun was tweeting "Have fun staying poor" at people who didn't fall for the latest iteration of the old get-rich-quick scam. That was how the whole idea of NFTs seems to be a great example of how clever people (or at least ones with specific smarts in areas like IT & cryptography) can also be really dumb. I'm reminded of the classic Larson cartoon of a geeky kid outside the Midvale School for the Gifted, stubbornly pushing at a door marked "pull".

My first thought about attempts to monetise a digital artworks by chaining it to a token of authenticity was how counter it runs to the principle that information wants to be free.* The legacy of some very smart digital pioneers is that reproducing digital information is trivially easy and almost costless. Attempting to make this process hard again is a difficult task which the smart people behind NFTs set themselves - and failed to achieve, as owning an NFT is not the same thing as owning the artwork or image, or text message, or tweet, or whatever else you decide to associated with it:

There is no possible way to see an NFT with your naked eyes. They are immaterial goods that you cannot see but own. NFTs are inherently treacherous and right-clickers, collectors, and artists worldwide are falling for their deception.

My second thought is how obviously mostly socially useless and scammy the NFT pioneers' project is. I say "mostly" because the quote above hints at how you could justify an attempt to make digital art, or any other digital creation, non-fungible. If you're an artist, or the creator of anything in the digital space, it would be easier to profit from your own hard work and talent if it wasn't possible for every rando on the internet to swipe your creation with a right click. If this was just a tool for creators to protect their creations from theft, I'd understand.

But it's not that. This is mostly middle men, trying to turn either someone else's work or some, usually ugly, mediocre, low-effort image they've created themselves into a something with the attributes of a gambling chip crossed with a share in a pyramid scheme which has value only if you can pursuade a horde of greedy and credulous people that it has value.

Other than that, I'll leave the commentary on this story to people who actually have a proper knowledge of IT, cryptography and finance, which I don't. But I think it's still legitimate even for me, as layperson, to take a firm view on this, based on the fact that there are plenty of explanations out there from crypto evangelists and from crypto sceptics who do have some background in this stuff. And I've found the arguments of the sceptics to be lucid where the evangelists are obscure, explanatory where the evangelists are defensive and disinterested, where the evangelists would have an obvious interest in pushing this stuff.

For an actually informed tear-down of NFT/crypto hype, explaining why this stuff mostly doesn't work as advertised (and would be a dystopian nightmare if implemented, even if it did work as advertised), see video below: 

*Here's the full orignal quote from Stewart Brand "On the one hand you have — the point you’re making Woz [Steve Wozniak] — is that information sort of wants to be expensive because it is so valuable — the right information in the right place just changes your life. On the other hand, information almost wants to be free because the costs of getting it out is getting lower and lower all of the time. So you have these two things fighting against each other."