net.wars

Appropriate privacy

At a workshop this week, one of the organizers posed a question that included the term “appropriate”. As in: “lawful access while maintaining appropriate user privacy”. We were there to think about approaches that could deliver better privacy and security over the next decade, with privacy defined as “the embedding of encryption or anonymization in software or devices”.

I had to ask: What work is “appropriate” doing in that sentence?

I had to ask because last weekend’s royal show was accompanied by preemptive arrests well before events began – at 7:30 AM. Most of the arrested were anti-monarchy protesters armed with luggage straps and placards, climate change protesters whose T-shirts said “Just Stop Oil”, and volunteers for the Night Stars on suspicion that the rape whistles they hand out to vulnerable women might be used to disrupt the parading horses. All of these had coordinated with the Metropolitan Police in advance or actually worked with them…which made no difference. All were held for many hours. Since then, the news has broken that an actual monarchist was arrested, DNA-sampled, fingerprinted, and held for 13 hours just for standing *near* some protesters.

It didn’t help the look of the thing that several days before the Big Show, the Met tweeted a warning that: “Our tolerance for any disruption, whether through protest or otherwise, will be low.”

The arrests were facilitated by the last-minute passage of the Public Order Act just days before with the goal of curbing “disruptive” protests. Among the now-banned practices is “locking on” – that is, locking oneself to a physical structure, a tactic the suffragettes used. among many others in campaigning for women’s right to vote. Because that right is now so thoroughly accepted, we tend to forget how radical and militant the Suffragists had to be to get their point across and how brutal the response was. A century from now, the mainstream may look back and marvel at the treatment meted out to climate change activists. We all know they’re *right*, whether or not we like their tactics.

Since the big event, the House of Lords has published its report on current legislation. The government is seeking to expand the Public Order Act even further by lowering the bar for “serious disruption” from “significant” and “prolonged” to “more than minor” and may include the cumulative impact of repeated protests in the same area. The House of Lords is unimpressed by these amendments via secondary legislation, first because of their nature, and second because they were rejected during the scrutiny of the original bill, which itself is only days old. Secondary legislation gets looked at less closely; the Lords suggest that using this route to bring back rejected provisions “raises possible constitutional issues”. All very Polite for accusing the government of abusing the system.

In the background, we’re into the fourth decade of the same argument between governments and technical experts over encryption. Technical experts by and large take the view that opening a hole for law enforcement access to encrypted content fatally compromises security; law enforcement by and large longs for the old days when they could implement a wiretap with a single phone call to a major national telephone company. One of the technical experts present at the workshop phrased all this gently by explaining that providing access enlarges the attack surface, and the security of such a system will always be weaker because there are more “moving parts”. Adding complexity always makes security harder.

This is, of course, a live issue because of the Online Safety bill, a sprawling mess of 262 pages that includes a requirement to scan public and private messaging for child sexual abuse material, whether or not the communications are encrypted.

None of this is the fault of the workshop we began with, which is part of a genuine attempt to find a way forward on a contentious topic, and whose organizers didn’t have any of this in mind when they chose their words. But hearing “appropriate” in that way at that particular moment raised flags: you can justify anything if the level of disruption that’s allowed to trigger action is vague and you’re allowed to use “on suspicion of” indiscriminately as an excuse. “Police can do what they want to us now,” George Monbiot writes at the Guardian of the impact of the bill.

Lost in the upset about the arrests was the Met’s decision to scan the crowds with live facial recognition. It’s impossible to overstate the impact of this technology. There will be no more recurring debates about ID cards because our faces will do the job. Nothing has been said about how the Met used it on the day, whether its use led to arrests (or on what grounds), or what the Met plans to do with the collected data. The police – and many private actors – have certainly inhaled the Silicon Valley ethos of “ask forgiveness, not permission”.

In this direction of travel, many things we have taken for granted as rights become privileges that can be withdrawn at will, and what used to be public spaces open to all become restricted like an airport or a small grocery store in Whitley Bay. This is the sliding scale in which “appropriate user privacy” may be defined.

Illustrations: Protesters at the coronation (by Alisdair Hickson at Wikimedia .

Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Follow on Mastodon or Twitter.

Strike two

Whatever happens with the Hollywood writers’ strike that began on Tuesday, the recent golden era of American TV, which arguably began with The Sopranos, is ending for viewers as well as creators.

A big reason for that golden era was that Hollywood’s loss of interest in grown-up movies pushed actors and writers who formerly looked down on TV to move across to where the more interesting work was finding a home. Another was the advent of streaming services, which competed with existing channels by offering creators greater freedom – and more money. It was never sustainable.

Streaming services’ business models are different. For nearly a decade, Netflix depended on massive debt to build a library to protect itself when the major studios ended their licensing deals. The company has so far gotten away with it because of (now ended) low interest rates and Wall Street’s focus on subscriber numbers in valuing its shares. Newer arrivals such as Amazon, Apple, and Disney can all finance loss-making startup streaming services from their existing businesses. All of these are members of the Alliance of Motion Picture and Television Producers, along with broadcast networks, cable providers, and motion picture studios. For the purposes of the strike, they are the “enemy”.

This landscape could not be more different than that of the last writers’ strike, in 2007-2008, when DVD royalties were important and streaming was the not-yet future. Of the technology companies refusing to bargain today, only Netflix was a player in 2007 – and it was then sending out DVDs by mail.

Essentially, what is happening to Hollywood writers is what happened to songwriters when music streaming services took over the music biz: income shrinkage. In 2021, veteran screenwriter Ken Levine, gave the detail of his persistently shrinking residuals (declining royalties paid for reuse). When American Airlines included an episode he directed of Everyone Loves Raymond in its transcontinental in-flight package for six months, his take from the thousands of airings was $1.19. He also documented, until he ended his blog in 2022, other ways writers are being squeezed; at Disconnect, Paris Marx provides a longer list. The Writers Guild of America’s declared goals are to redress these losses and bring residuals and other pay on streaming services into line with older broadcasters.

Even an outsider can see the bigger picture: broadcast networks, traditionally the biggest payers, are watching their audiences shrink and retrenching, and cable and streaming services commission shorter seasons, which they renew at a far more leisurely pace. Also a factor is the shift in which broadcast networks reair their new shows a day or two later on their streaming service. The DVD royalties that mattered in the 2007-2008 strike are dying away, and just as in music royalties from streaming are a fraction of the amount. Overall, the WGA says that in the last decade writers’ average incomes have dropped by 4% – 23% if you include inflation. Meanwhile, industry profits have continued to rise.

The new issue on the block is AI – not because large language models are good enough to generate good scripts (as if), but because writers fear the studios will use them to generate crappy scripts and then demand that the writers rewrite them into quality for a pittance. Freelance journalists have already reported seeing publishers try this gambit.

In 2007, 2007, and again in 2017, Levine noted that the studios control the situation. They can make a deal and end the strike any time they decide it’s getting too expensive or disruptive. Eventually, he said, the AMPTP will cut a deal, writers will get some of what they need, and everyone will go back to work. Until then, the collateral damage will mount to writers and staff in adjacent industries and California’s economy. At Business Insider, Lucia Moses suggests that Netflix, Amazon, and Disney all have enough content stockpiled to see them through.

Longer-term, there will be less predictable consequences. In 2007-2008, Leigh Blickley reported in a ten-years-later lookback at the Huffington Post, these included the boom in “unscripted” reality TV and the death of pathways into the business for new writers.

Underlying all this is a simple but fundamental change. Broadcast networks cared what Americans watched because their revenues depended on attracting large audiences that advertisers would pay to reach. Until VCRs arrived to liberate us from the tyranny of schedules, the networks competed on the quality and appeal of their programming in each time slot. Streaming services compete on their whole catalogue, and care only that you subscribe; ratings don’t count.

The WGA warns that the studios’ long-term goal is to turn screenwriting into gig economy work. In 2019, at BIG, Matt Stoller warned that Netflix was predatorily killing Hollywood, first by using debt financing to corner the market, and second by vertically integrating its operation. Like the the studios that were forced to divest their movie theaters in 1948, Netflix, Amazon, and Apple own content, controls its distribution, and sells retail access. It should be no surprise if a vertically integrated industry with a handful of monopolistic players cuts costs by treating writers the way Uber treats drivers: enshittification.

The WGA’s 12,000 members know their skills, which underpin a trillion-dollar industry, are rare. They have a strong union and a long history of solidarity. If they can’t win against modern corporate extraction, what hope for the rest of us?

Illustrations: WGA members picketing in 2007 (by jengod at Wikimedia).

The privacy price of food insecurity

One of the great unsolved questions continues to be: what is my data worth? Context is always needed: worth to whom, under what circumstances, for what purpose? Still, supermarkets may give us a clue.

At Novara Media, Jake Hurfurt, who runs investigations for Big Brother Watch, has been studying suprmarket loyalty cards. He finds that increasingly only loyalty card holders have access to special offers, which used to be open to any passing customer.

Tesco now and Sainsburys soon, he says, “are turning the cost-of-living crisis into a cost-of-privacy crisis”,

Neat phrasing, but I’d say it differently: these retailers are taking advantage of the cost-of-living crisis to extort desperate people ito giving up their data. The average value of the discounts might – for now – give a clue to the value supermarkets place on it.

But not for long, since the pattern going forward is a predictable one of monopoly power: as the remaining supermarkets follow suit and smaller independent shops thin out under the weight of rising fuel bills and shrinking margins, and people have fewer choices, the savings from the loyalty card-only special offers will shrink. Not so much that they won’t be worth having, but it seems obvious they’ll be more generous with the discounts – if “generous” is the word – in the sign-up phase than they will once they’ve achieved customer lock-in.

The question few shoppers are in a position to answer while they’re strying to lower the cost of filling their shopping carts is what the companies do with the data they collect. BBW took the time to analyze Tesco’s and Sainsburys’ privacy policies, and found that besides identity data they collect detailed purchase histories as well as bank accounts and payment information…which they share with “retail partners, media partners, and service providers”. In Tesco’s case, these include Facebook, Google, and, for those who subscribe to them, Virgin Media and Sky. Hyper-targeted personal ads right there on your screen!

All that sounds creepy enough. But consider what could well come next. Also this week, a cross-party group of 50 MPs and peers and cosinged by BBW, Privacy International and Liberty, wrote to Frasers Group deploring that company’s use of live facial recognition in its stores, which include Sports Direct and the department store chain House of Fraser. Frasers Group’s purpose, like retailers and pub chains were trialing a decade ago , is effectively to keep out people suspected of shoplifting and bad behavior. Note that’s “suspected”, not “convicted”.

What happens as these different privacy invasions start to combine?

A store equipped with your personal shopping history and financial identity plus live facial recognition cameras, knows the instant you walk into the store who you are, what you like to buy, and how valuable a customer your are. Such a system, equipped with some sort of socring, could make very fine judgments. Such as: this customer is suspected of stealing another customer’s handbag, but they’re highly profitable to us, so we’ll let that go. Or: this customer isn’t suspected of anything much but they look scruffy and although they browse they never buy anything – eject! Or even: this journalist wrote a story attacking our company. Show them the most expensive personalized prices. One US entertainment company is already using live facial recognition to bar entry to its venues to anyone who works for any law firm involved in litigation against it. Britain’s data protection laws should protect us against that sort of abuse, but will they survive the upcoming bonfire of retained EU law?

And, of course, what starts with relatively anodyne product advertising becomes a whole lot more sinister when it starts getting applied to politics, voter manipulation and segmentation, and the “pre-crime” systems

Add the possibilities of technology that allows retailers to display personalized pricing in-store, just like an online retailer could do in the privacy of your own browser, Could we get to a scenario where a retailer, able to link your real world identity and purchasing power to your online nd offline movements could perform a detailed calculation of what you’d be willing to pay for a particular item? What would surge pricing for the last remaining stock of the year’s hottest toy on Christmas Eve look like?

This idea allows me to imagine shopping partnerships, where the members compare prices and the partner with the cheapest prices buys that item for the whole group. In this dystopian future, I imagine such gambits would be banned.

Most of this won’t affect people rich enough to grandly refuse to sign up for loyalty cards, and none of it will affect people rich and eccentric enough to do source everything from local, independent shops – and, if they’re allowed, pay cash.

Four years ago, Jaron Lanier toured with the proposal that we should be paid for contributing to commercial social media sites. The problem with this idea was and is that payment creates a perverse incentive for users to violate their own privacy even more than they do already, and that fair payment can’t be calculated when the consequences of disclosure are perforce unknown.

The supermarket situation is no different. People need food security and affordability, They should not have to pay for that with their privacy.

Illustrations: .London supermarket checkout, 2006 (via Wikimedia.

Breaking badly

This week, the Online Safety Bill reached the House of Lords, which will consider 300 amendments. There are lots of problems with this bill, but the one that continues to have the most campaigning focus is the age-old threat to require access to end-to-end encrypted messaging services.

At his blog, security consultant Alec Muffett predicts the bill will fail in implementation if it passes. For one thing, he cites the argument made by Richard Allan, Baron of Hallam that the UK government wants the power to order decryption but will likely only ever use it as a threat to force the technology companies to provide other useful data. Meanwhile, the technology companies have pushed back with an open letter saying they will withdraw their encrypted products from the UK market rather than weaken them.

In addition, Muffett believes the legally required secrecy when a service provider is issued with a Technical Capability Notice to provide access to communications, which was devised for the legacy telecommunications world, is impossible in today’s world of computers and smartphones. Secrecy is no longer possible, given the many researchers and hackers who make it their job to study changes to apps, and who would surely notice and publicize new decryption capabilities. The government will be left with the choice of alienating the public or failing to deliver its stated objectives.

At Computer Weekly, Bill Goodwin points out that undermining encryption will affect anyone communicating with anyone in Britain, including the Ukrainian military communicating with the UK’s Ministry of Defence.

Meanwhile, this week Ed Caesar reports at The New Yorker on law enforcement’s successful efforts to penetrate communications networks protected by Encrochat and Sky ECC. It’s a reminder that there are other choices besides opening up an entire nation’s communications to attack.

***

This week also saw the disappointing damp-squib settlement of the lawsuit brought by Dominion Voting Systems against Fox News. Disappointing, because it leaves Fox and its hosts free to go on wreaking daily havoc across America by selling their audience rage-enhanced lies without even an apology. The payment that Fox has agreed to – $787 million – sounds like a lot, but a) the company can afford it given the size of its cash pile, and b) most of it will likely be covered by insurance.

If Fox’s major source of revenues were advertising, these defamation cases – still to come is a similar case brought by Smartmatic – might make their mark by alienating advertisers, as has been happening with Twitter. But it’s not; instead, Fox is supported by the fees cable companies pay to carry the channel. Even subscribers who never watch it are paying monthly for Fox News to go on fomenting discord and spreading disinformation. And Fox is seeking a raise to $3 per subscriber, which would mean more than $1,8 billion a year just from affiliate revenue.

All of that insulates the company from boycotts, alienated advertisers, and even the next tranche of lawsuits. The only feedback loop in play is ratings – and Fox News remains the most-watched basic cable network.

This system could not be more broken.

***

Meanwhile, an era is ending: Netflix will mail out its last rental DVD in September. As Chris Stokel-Walker writes at Wired, the result will be to shrink the range of content available by tens of thousands of titles because the streaming library is a fraction of the size of the rental library.

This reality seems backwards. Surely streaming services ought to have the most complete libraries. But licensing and lockups mean that Netflix can only host for streaming what content owners decree it may, whereas with the mail rental service once Netflix had paid the commercial rental rate to buy the DVD it could stay in the catalogue until the disk wore out.

The upshot is yet another data point that makes pirate services more attractive: no ads, easy access to the widest range of content, and no licensing deals to get in the way.

***

In all the professions people have been suggesting are threatened by large language model-based text generation – journalism, in particular – no one to date has listed fraudulent spiritualist mediums. And yet…

The family of Michael Schumacher is preparing legal action against the German weekly Die Aktuelle for publishing an interview with the seven-time Formula 1 champion. Schumacher has been out of the public eye since suffering a brain injury while skiing in 2013. The “interview” is wholly fictitious, the quotes created by prompting an “AI” chat bot.

Given my history as a skeptic, my instinctive reaction was to flash on articles in which mediums produced supposed quotes from dead people, all of which tended to be anodyne representations bereft of personality. Dressing this up in the trappings of “AI” makes such fakery no less reprehensible.

An article in the Washington Post examines Google’s C4 data set scraped from 15 million websites and used to train several of the highest profile large language models. The Post has provided a search engine, which tells us that my own pelicancrossing.net, which was first set up in 1996, has contributed 160,000 words or phrases (“tokens”), or 0.0001% of the total. The obvious implication is that LLM-generated fake interviews with famous people can draw on things they’ve actually said in the past, mixing falsity and truth into a wasteland that will be difficult to parse.

Illustrations: The House of Lords in 2011 (via Wikimedia).

Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Follow on Twitter.

Ex libris

So as previously discussed here three years ago and two years ago, on March 24 the US District Court for the Southern District of New York found that the Internet Archive’s controlled digital lending fails copyright law. Half of my social media feed on this subject filled immediately with people warning that publishers want to kill libraries and this judgment is a dangerous step limiting access to information; the other half is going “They’re stealing from authors. Copyright!” Both of these things can be true. And incomplete.

To recap: in 2006 the Internet Archive set up the Open Library to offer access to digitized books under “controlled digital lending”. The system allows each book to be “out” on “loan” to only one person at a time, with waiting lists for popular titles. In a white paper, lawyers David R. Hansen and Kyle K. Courtney call this “format shifting” and say that because the system replicates library lending it is fair use. Also germane: the Archive points to a 2007 California decision that it is in fact a library. Other countries may beg to differ.

When public libraries closed at the beginning of the covid19 pandemic, the Internet Archive announced the National Emergency Library, which suspended the one-copy-at-a-time rule and scrubbed the waiting lists so anyone could borrow any book at any time. The resulting publicity was the first time many people had heard of the Open Library, although authors had already complained. Hachette Book Group, Penguin Random House, HarperCollins, and Wiley filed suit. Shortly afterwards, the Archive shut down the National Emergency Library. The Open Library continues, and the Archive will appeal the judge’s ruling.

On the they’re-killing-the-libraries side: Mike Masnick and Fight for the Future. At Walled Culture, Glyn Moody argues that sharing ebooks helps sell paid copies. Many authors agree with the publishers that their living is at risk; a group of exceptions including Neil Gaiman, Naomi Klein, and Cory Doctorow, have published an open letter defending the Archive.

At Vice, Claire Woodstock lays out some of the economics of library ebook licenses, which eat up budgets but leave libraries vulnerable and empty-shelved when a service is withdrawn. She also notes that the Internet Archive digitizes physical copies it buys or receives as donations, and does not pay for ebook licenses.

Brief digression back to 1996, when Pamela Samuelson warned of the coming copyright battles in Wired. Many of its key points have since either been enshrined into law, such as circumventing copy protection; others, such as requiring Internet Service Providers to prevent users from uploading copyrighted material, remain in play today. Number three on her copyright maximalists’ wish listeliminating first-sale rights for digitally transmitted documents. This is the doctrine that enables libraries to lend books.

It is therefore entirely believable that commercial publishers believe that every library loan is a missed sale. Outside the US, many countries have a public lending right that pays royalties on loans for that sort of reason. The Internet Archive doesn’t pay those, either.

It surely isn’t facing the headwinds public libraries are. In the UK, years of austerity have shrunk library budgets and therefore their numbers and opening hours. In the US, libraries are fighting against book bans; in Missouri, the Republican-controlled legislature voted to defund the state’s libraries entirely, apparently in retaliation.

At her blog, librarian and consultant Karen Coyle, who has thought for decades about the future of libraries, takes three postings to consider the case. First, she offers a backgrounder, agreeing that the Archive’s losing on appeal could bring consequences for other libraries’ digital lending. In the second, she teases out the differences between academic/research libraries and public libraries and between research and reading. While journals and research materials are generally available in electronic format, centuries of books are not, and scanned books (like those the Archive offers) are a poor reading experience compared to modern publisher-created ebooks. These distinctions are crucial to her third posting, which traces the origins of controlled digital lending.

As initially conceived by Michelle M. Wu in a 2011 paper for Law Library Journal, controlled digital lending was a suggestion that law libraries could, either singly or in groups, buy a hard copy for their holdings and then circulate a digitized copy, similar to an Inter-Library Loan. Law libraries serve limited communities, and their comparatively modest holdings have a known but limited market.

By contrast, the Archive gives global access to millions of books it has scanned. In court, it argued that the availability of popular commercial books on its site has not harmed publishers’ revenues. The judge disagreed: the “alleged benefits” of access could not outweigh the market harm to the four publishers who brought the suit. This view entirely devalues the societal role libraries play, and Coyle, like many others, is dismayed that the judge saw the case purely in terms of its effect on the commercial market.

The question I’m left with is this: is the Open Library a library or a disruptor? If these were businesses, it would obviously be the latter: it avoids many of the costs of local competitors, and asks forgiveness not permission. As things are, it seems to be both: it’s a library for users, but a disruptor to some publishers, some authors, and potentially the world’s libraries. The judge’s ruling captures none of this nuance.

Illustrations: 19th century rendering of the Great Library of Alexandria (via Wikimedia).

Review: Cloudmoney

Cloudmoney: Cash, Cards, Crypto, and the War for our Wallets
By Brett Scott
Publisher: Bodley Head
ISBN: 978-1-847-92587-9

Three years ago, the area around the local tube station included a bank and four ATMs. Come the pandemic, the bank closed, never to return, and so did two of the ATMs. The loss of the bank gave a couple of the chain stores an excuse to refuse to take cash. But they’re a minority in an area full of independent local shops, who recognize that many of their customers are cash users. Journey into some parts of central London, however, and cash gets you ghosted.

We are told that the cashless future is what we want: it’s more convenient (except when the system is down, the app needs to be rebooted, or there’s no Internet connection). The reality, as “monetary anthropologist” and former broker Brett Scott points out in his book Cloudmoney, is that despite this inevitability narrative, one reason electronic/digital payments are more convenient is a deliberate effort to make cash harder to access. Often, promoters claim the cashless society is – or will be – more financially inclusive. Yet, as Scott recounts, that “inclusion” in the remote global economy often brings with it the exclusion of locally-controlled, less formal economies. Less financial inclusion, more *enclosure* and “corporate seep”.

Scott’s central thesis is simple: once the forces of Big Tech and Big Finance have merged, they will have a hitherto unimaginable amount of power over all of us. I have some sympathy with this argument. People forget that it was through the banks that Gilead was brought into being in Margaret Atwood’s The Handmaid’s Tale. All they had to do was locate all the accounts tagged “F” and turn off access until a suitable male came forward to claim them. This is the power of cloudmoney – money that exists for us only in the form of numbers that represent promises to pay. Scott is not predicting a specific dystopia; but he does want to propagate a counterbalancing narrative to the “liberation” every new fintech app pretends to promise while scarfing up all our personal data. In his campaign to protect the public system of cash, he sometimes finds himself in the company of conspiracy theorists whose other ideas he rejects.

What is less clear is where bitcoin and other cryptocurrencies fit in. They also started with rhetoric: they were digital cash, digital gold, a mechanism for bypassing the world’s banks and governments. In practice, so far, they haven’t succeeded at any of these things, and even in El Salvador, where bitcoin is legal tender, you can’t use it to buy a box of oatmeal in a supermarket.

The story technology companies tell is, of course, that they are disrupting the stodgy, antiquated world of traditional finance. Instead, what Scott sees is plain old automation that serves that world and tightens its control. Almost every new service, whatever the rhetoric it starts with, from credit cards to Paypal to Apple Pay to Facebook’s failed Libra cryptocurrency, becomes a front end for bank accounts for the same reason that robbers always focused on them: that’s where the money is. The exception is cash – slow, partially disconnected cash that enables transactions that aren’t caught in what Scott calls the “digital mesh” of corporate capitalism. No wonder they hate it.

Excluding the vote

“You have to register at home, where your parents live,” said the clerk at the Board of Elections office.

I was 18, and registering to vote for the first time. It was 1972.

“I don’t live there,” I said. “I live here.” “Here” was Ithaca, NY, a town that, I learned later, was hyper-conscious that college students – Cornell, Ithaca College – outnumbered local residents. They didn’t want us interlopers overwhelming their preferences.

We had a couple more back-and-forths like this, and then she picked up the phone and called the state authorities in Albany for an official ruling. I knew – or thought I knew – that the law was on my side.

It was. I registered. I voted.

In about a month, the UK will hold local elections. For the first time, anyone presenting themselves to vote at the polls will be required to show an ID card with a photograph. This is a policy purely imported from American Republicans, and it has no basis in necessity. The Electoral Commission, in recommending its introduction, admitted that the issue was public perception. The big issues with respect to elections are around dark money and the processes by which candidates are chosen.

For 49 days in the fall of 2022, Liz Truss served as prime minister; she was chosen by 81,326 Tory party members. Out of the country’s roughly 68 million people, only 141,725 (out of an estimated 172,000 party members) voted in that contest because, since the Conservatives had decisively won the 2019 election, they were just electing a new leader. Rishi Sunak was voted in by 202 MPs.

The government’s proximate excuse for bringing in voter ID is the fraud-riddled May 2014 mayoral election in the London borough of Tower Hamlets. Four local residents risked their own money to challenge the outcome, and in 2015 won an Election Court ruling voiding the election and barring the cheating winner from standing for public office for five years. Their complaints; included vote-rigging, false statements made by the winning candidates about his rival, bribery, and religious influence.

The High Court of Justice’s judgment in the case says: “…in practice, where electoral malpractice is established, particularly in the field of vote-rigging, it is very rare indeed to find members of the general public engaging in DIY vote-rigging on behalf of a candidate. Generally speaking, if there is widespread personation or false registration or misuse of postal votes, it will have been organised by the candidate or by someone who is, in law, his agent.”

Surely a more logical response to the Tower Hamlets case would be to make it easier – or at least quicker – for individuals to challenge election results and examine ways to ensure better behavior by *candidates*, not voters.

The judgment also notes that personation – assuming someone else’s identity in order to vote – was far more of a risk when fewer people qualified to vote. There followed a long period when it was too labor-intensive for too little reward; you need a lot of impersonators to change the result. In recent years, however, postal voting has made it viable again; in two wards of a 2008 Birmingham election Labour candidates committed 15 types of fraud involving postal ballots. The election in those two wards was re-run.

In his book Security Engineering, Cambridge professor Ross Anderson notes that the likelihood that expanded use of postal ballots would open the way for vote-buying an intimidation was predicted even as first Margaret Thatcher and then Tony Blair pursued the policy. But the main point is clear: the big problem is postal ballots, which you can’t solve by requiring voter ID from those who vote in person. It’s the wrong threat model. As Anderson observes, “…it’s typically the incumbent who tweaks the laws, buys the voting machines, and creates as many advantages for their own side, small and large, as the local political culture will tolerate.”

But voter ID is the policy that Boris Johnson used his 80-seat majority to push through in the form of the Elections Act (2022), which also weakens the independence of the Electoral Commission. As the bill went through Parliament, estimates were that about 3.5 million people lacked any qualifying form of ID, and that those 3.5 million skew heavily toward people who are not expected to vote Conservative.

This was all maddening enough – and then they published the list of acceptable forms of ID. Tl;dr: the list blatantly skews in favor of older and richer people, who are presumed to be more likely to vote Conservative. Passports, driving licenses, and travel passes 60+ for people are all acceptable. Student ID cards and travel cards and passesare not. The government says they are not secure enough, a bit like saying a lock on the door is pointless because it’s not a burglar alarm.

There is a scheme for issuing free voter cards; applications must be in by April 25. People can also vote by post or by proxy without ID. And there are third parties pushing paid ID cards, too. But what it comes down to is next month a bunch of people are going to go to vote and will be barred. And this from the same people who wanted online voting to “increase access”.

Illustrations: London polling station 2017 (by Mramoeba at Wikimedia.

Unclear and unpresent dangers

Monthly computer magazines used to fret that their news pages would be out of date by the time the new issue reached readers. This week in AI, a blog posting is out of date before you hit send.

This – Friday – morning, the Italian data protection authority, Il Garante, has ordered ChatGPT to stop processing the data of Italian users until it complies with the General Data Protection Regulation. Il Garante’s objections, per Apple’s translation, posted by Ian Brown: ChatGPT provides no legal basis for collecting and processing its massive store of the personal data used to train the model, and that it fails to filter out users under 13.

This may be the best possible answer to the complaint I’d been writing below.

On Wednesday, the Future of Life Institute published an open letter calling for a six-month pause on developing systems more powerful than Open AI’s current state of the art, GPT4. Barring Elon Musk, Steve Wozniack, and Skype co-founder Jaan Tallinn, most of the signatories are unfamiliar names to most of us, though the companies and institutions they represent aren’t – Pinterest, the MIT Center for Artificial Intelligence, UC Santa Cruz, Ripple, ABN-Amro Bank. Almost immediately, there was a dispute over the validity of the signatures..

My first reaction was on the order of: huh? The signatories are largely people who are inventing this stuff. They don’t have to issue a call. They can just *stop*, work to constrain the negative impacts of the services they provide, and lead by example. Or isn’t that sufficiently performative?

A second reaction: what about all those AI ethics teams that Silicon Valley companies are disbanding? Just in the last few weeks, these teams have been axed or cut at Microsoft and Twitch; Twitter of course ditched such fripperies last November in Musk’s inaugural wave of cost-cutting. The letter does not call to reinstate these.

The problem, as familiar critics such as Emily Bender pointed out almost immediately, is that the threats the letter focuses on are distant not-even-thunder. As she went on to say in a Twitter thread, the artificial general intelligence of the Singularitarian’s rapture is nowhere in sight. By focusing on distant threats – longtermism – we ignore the real and present problems whose roots are being continuously more deeply embedded into the new-building infrastructure: exploited workers, culturally appropriated data, lack of transparency around the models and algorithms used to build these systems….basically, all the ways they impinge upon human rights.

This isn’t the first time such a letter has been written and circulated. In 2015, Stephen Hawking, Musk, and about 150 others similarly warned of the dangers of the rise of “superintelligences”. Just a year later, in 2016, Pro Publica investigated the algorithm behind COMPAS, a risk-scoring criminal justice system in use in US courts in several states. Under Julia Angwin‘s scrutiny, the algorithm failed at both accuracy and fairness; it was heavily racially biased. *That*, not some distant fantasy, was the real threat to society.

“Threat” is the key issue here. This is, at heart, a letter about a security issue, and solutions to security issues are – or should be – responses to threat models. What is *this* threat model, and what level of resources to counter it does it justify?

Today, I’m far more worried by the release onto public roads of Teslas running Full Self Drive helmed by drivers with an inflated sense of the technology’s reliability than I am about all of human work being wiped away any time soon. This matters because, as Jessie Singal, author of There Are No Accidents, keeps reminding us, what we call “accidents” are the results of policy decisions. If we ignore the problems we are presently building in favor of fretting about a projected fantasy future, that, too, is a policy decision, and the collateral damage is not an accident. Can’t we do both? I imagine people saying. Yes. But only if we *do* both.

In a talk this week for a group at the French international research group AI Act. This effort began well before today’s generative tools exploded into public consciousness, and isn’t likely to conclude before 2024. It is, therefore, much more focused on the kinds of risks attached to public sector scandals like COMPAS and those documented in Cathy O’Neil’s 2017 book Weapons of Math Destruction, which laid bare the problems with algorithmic scoring with little to tether it to reality.

With or without a moratorium, what will “AI” look like in 2024? It has changed out of recognition just since the last draft text was published. Prediction from this biological supremacist: it still won’t be sentient.

All this said, as Edwards noted, even if the letter’s proposal is self-serving, a moratorium on development is not necessarily a bad idea. It’s just that if the risk is long-term and existential, what will six months do? If the real risk is the hidden continued centralization of data and power, then those six months could be genuinely destructive. So far, it seems like its major function is as a distraction. Resist.

Illustrations: IBM’s Watson, which beat two of Jeopardy‘s greatest champions in 2011. It has since failed to transform health care.

Memex 2.0

As language models get cheaper, it’s dawned on me what kind of “AI” I’d like to have: a fully personalized chat bot that has been trained on my 30-plus years of output plus all the material I’ve read, watched, listened to, and taken notes on all these years. A clone of my brain, basically, with more complete and accurate memory updated alongside my own. Then I could discuss with it: what’s interesting to write about for this week’s net.wars?

I was thinking of what’s happened with voice synthesis. In 2011, it took the Scottish company Cereproc months to build a text-to-speech synthesizer from recordings of Roger Ebert’s voice. Today, voice synthesizers are all over the place – not personalized like Ebert’s, but able to read a set text plausibly enough to scare voice actors.

I was also thinking of the Stochastic Parrots paper, whose first anniversary was celebrated last week by authors Emily Bender, Timnit Gebru, Angelina McMillan-Major, and Margaret Mitchell. An important part of the paper advocates for smaller, better-curated language models: more is not always better. I can’t find a stream for the event, but here’s the reading list collected during the proceedings. There’s lots I’d rather eliminate from my personal assistant. Eliminating unwanted options upfront has long been a widspread Internet failure, from shopping sites (“never show me pet items”) to news sites (“never show me fashion trends”). But that sort of selective display is more difficult and expensive than including everything and offering only inclusion filters.

A computational linguistics expert tells me that we’re an unknown amount of time away from my dream of the wg-bot. Probably, if such a thing becomes possible it will be based on someone’s large language model and fine-tuned with my stuff. Not sure I entirely like this idea; it means the model will be trained on stuff I haven’t chosen or vetted and whose source material is unknown, unless we get a grip on forcing disclosure or the proposed BLOOM academic open source language model takes over the world.

I want to say that one advantage to training a chatbot on your own output is you don’t have to worry so much about copyright. However, the reality is that most working writers have sold all rights to most of their work to large publishers, which means that such a system is a new version of digital cholera. In my own case, by the time I’d been in this business for 15 years, more than half of the publications I’d written for were defunct. I was lucky enough to retain at least non-exclusive rights to my most interesting work, but after so many closures and sales I couldn’t begin to guess – or even know how to find out – who owns the rights to the rest of it. The question is moot in any case: unless I choose to put those group reviews of Lotus 1-2-3 books back online, probably no one else will, and if I do no one will care.

On Mastodon, the specter of the upcoming new! improved! version of the copyright wars launched by the arrival of the Internet: “The real generative AI copyright wars aren’t going to be these tiny skirmishes over artists and Stability AI. Its going to be a war that puts filesharing 2.0 and the link tax rolled into one in the shade.” Edwards is referring to this case, in which artists are demanding billions from the company behind the Stable Diffusion engine.

Edwards went on to cite a Wall Street Journal piece that discusses publishers’ alarmed response to what they perceive as new threats to their business. First: that the large piles of data used to train generative “AI” models are appropriated without compensation. This is the steroid-fueled analogue to the link tax, under which search engines in Australia pay newspapers (primarily the Murdoch press) for including them in news search results. A similar proposal is pending in Canada.

The second is that users, satisfied with the answers they receive from these souped-up search services will no longer bother to visit the sources – especially since few, most notably Google, seem inclined to offer citations to back up any of the things they say.

The third is outright plagiarism without credit by the chatbot’s output, which is already happening.

The fourth point of contention is whether the results of generative AI should be themselves subject to copyright. So far, the consensus appears to be no, when it comes to artwork. But some publishers who have begun using generative chatbots to create “content” no doubt claim copyright in the results. It might make more sense to copyright the *prompt*. (And some bright corporate non-soul may yet try.)

At Walled Culture, Glyn Moody discovers that the EU has unexpectedly done something right by requiring positive opt-in to copyright protection against text and data mining. I’d like to see this as a ray of hope for avoiding the worst copyright conflicts, but given the transatlantic rhetoric around privacy laws and data flows, it seems much more likely to incite another trade conflict.

It now dawns on me that the system I outlined in the first paragraph is in fact Vannevar Bush’s Memex. Not the web, which was never sufficiently curated, but this, primed full of personal intellectual history. The “AI” represents those thousands of curating secretaries he thought the future would hold. As if.

Illustrations: Stable Diffusion rendering of “stochastic parrots”, as prompted by Jon Crowcroft.

Performing intelligence

“Oh, great,” I thought when news broke of the release of GPT-4. “Higher-quality deception.”

Most of the Internet disagreed; having gone mad only a few weeks ago over ChatGPT, everyone’s now agog over this latest model. It passed all these tests!

One exception was the journalist Paris Marx, who commented on Twitter: “It’s so funny to me that the AI people think it’s impressive when their programs pass a test after being trained on all the answers.”

Agreed. It’s also so funny to me that they call that “AI” and don’t like it when researchers like computational linguist Emily Bender call it a “stochastic parrot”. At Marx’s Tech Won’t Save Us podcast, Goldsmith professor Dan McQuillan, author of Resisting AI: An Anti-fascist Approach to Artificial Intelligence, calls it a “bullshit engine” whose developers’ sole goal is plausibility – plausibility that, as Bender has said, allows us imaginative humans to think we detect a mind behind it, and the result is to risk devaluing humans.

Let’s walk back to an earlier type of system that has been widely deployed: benefits scoring systems. A couple of weeks ago, Lighthouse Reports and Wired magazine teamed up on an investigation of these systems, calling them “suspicion machines”.

Their work focuses on the welfare benefits system in use in Rotterdam between 2017 and 2021, which uses 315 variables to risk-score benefits recipients according to the likelihood that their claims are fraudulent. In detailed, worked case analyses, they find systemic discrimination: you lose points for being female, for being female and having children (males aren’t asked about children), for being non-white, and for ethnicity (knowing Dutch a requirement for welfare recipients). Other variables include missing meetings, age, and “lacks organizing skills”, which was just one of 54 variables based on case workers’ subjective assessments. Any comment a caseworker adds translates to a 1 added to the risk score, even if it’s positive. The top-scoring 10% are flagged for further investigation.

This is the system that Accenture, the city’s technology partner on the early versions, said at its unveiling in 2018 was an “ethical solution” and promised “unbiased citizen outcomes”. Instead, Wired says, the algorithm “fails the city’s own test of fairness”.

The project’s point wasn’t to pick on Rotterdam; of the dozens of cities they contacted it just happened to be the only one that was willing to share the code behind the algorithm, along with the list of variables, prior evaluations, and the data scientists’ handbook. It even – after being threatened with court action under freedom of information laws, shared the mathematical model itself.

The overall conclusion: the system was so inaccurate it was little better than random sampling “according to some metrics”.

What strikes me, aside from the details of this design, is the initial choice of scoring benefits recipients for risk of fraud. Why not score them for risk of missing out on help they’re entitled to? The UK government’s figures on benefits fraud indicate that in 2021-2022 overpayment (including error as well as fraud) amounted to 4%; and *underpayment* 1.2% of total expenditure. Underpayment is a lot less, but it’s still substantial (£2.6 billion). Yes, I know, the point of the scoring system is to save money, but the point of the *benefits* system is to help people who need it. The suspicion was always there, but the technology has altered the balance.

This was the point the writer Ellen Ullman noted in her 1996 book Close to the Machine”: the hard-edged nature of these systems and their ability to surveil people in new ways, “infect” their owners with suspicion even of people they’ve long trusted and even when the system itself was intended to be helpful. On a societal scale, these “suspicion machines” embed increased division in our infrastructure; in his book, McQuillan warns us to watch for “functionality that contributes to violent separations of ‘us and them’.”

Along those lines, it’s disturbing that Open AI, the owner of ChatGPT and GPT-4 (and several other generative AI gewgaws) has now decided to keep secret the details of its large language models. That is, we have no sight into what data was used in training, what software and hardware methods were used, or how energy-intensive it is. If there’s a machine loose in the world’s computer systems pretending to be human, shouldn’t we understand how it works? It would help with damping down imagining we see a mind in there.

The company’s argument appears to be that because these models could become harmful it’s bad to publish how they work because then bad actors will use them to create harm. In the cybersecurity field we call this “security by obscurity” and there is a general consensus that it does not work as a protection.

In a lengthy article at New York magazine, Elizabeth Weil. quotes Daniel Dennett’s assessment of these machines: “counterfeit people” that should be seen as the same sort of danger to our system as counterfeit money. Bender suggests that rather than trying to make fake people we should be focusing more on making tools to help people.

The thing that makes me tie it to the large language models that are producing GPT is that in both cases it’s all about mining our shared cultural history, with all its flaws and misjudgments, in response to a prompt and pretending the results have meaning and create new knowledge. And *that’s* what’s being embedded into the world’s infrastructure. Have we learned nothing from Clever Hans?

Illustrations: Clever Hans, performing in Leipzig in 1912 (by Karl Krali, via Wikimedia.