Slow down AI?

An open letter, hosted by the Future of Life Institute, calling for a pause in the development of Artificial Intelligence has now been signed by over 1,850 people, including Steve Wozniak and Elon Musk* (who funds the Future of Life Institute). That number may be less impressive than at first might seem — if you log on to the FOLI site (link above) you will be invited to sign the letter yourself. OK, warnings about AI may be needed — would that they were heeded. I seem to remember Stephen Hawking giving us a similar warning a few years ago — turns out it was in 2014 when he told the BBC, “The development of full artificial intelligence could spell the end of the human race.”

Sigal Samuel has an article at Vox, discussing the idea that for all our sakes we should slow down research into AI. She herself is in favor of a pause in research but identifies three problems which people may have with this idea:

• Objection 1: “Technological progress is inevitable, and trying to slow it down is futile”
• Objection 2: “We don’t want to lose an AI arms race with China”
• Objection 3: “We need to play with advanced AI to figure out how to make advanced AI safe”

Not much of an argument against “the end of the human race” I fear. On an idiot level the objection to Ms Samuel’s objections is that her whole discussion is too USA-focussed. Let’s imagine that the US Congress did pass a law mandating a halt in AI research (which obviously isn’t going to happen) and the State Department were to persuade China to do likewise (which I suspect would be less unlikely), what would that achieve? Is such a thing at all enforceable, even within the USA? And last time I looked there were other nations in the world, and many of them seem to have access to computers! And the unfortunate reality is that research is carried out by people not nations, and undercover computer activities may be assumed to continue. Of course funding might be sparser, and there would no doubt be a sort of shaming influence, or the other way round, the urge to follow a good example, but I expect lots of people in lots of places would be perfectly happy, nay even eager, since now without competition, to forge ahead. Ms Samuel refers to technologies we have managed to “halt” including human cloning. OK, we know that the guy who did it was arrested, but to the ambitious human engineer that is probably more of a lesson on how to be more discrete than a warning to stop.

A pause would however be delightful. We hear too much about all this these days. But we do need time to think about what we want the copyright law to look like now that there are such things as computers that can do such things as they now can. Just saying “No” doesn’t really amount to a policy. This may be a somewhat secondary issue for most of the world, but copyright is the bread and butter of media businesses. I hope someone in Washington is thinking about copyright reform, or at least thinking that thinking about it might be something they should get round to soon.

Once the genie is out of the bottle, we know, don’t we, that it can’t be put back. Every new technology has been greeted with foreboding — think of Socrates and the supposed baleful influence of writing stuff down in books. This is not to disagree with Stephen Hawking — after all, he does say “could spell the end” not “will”. As far as I can see we really have no alternative to stumbling forward and keeping our fingers crossed. Maybe we’ll get lucky, and just as the breakneck speed of a fifty-mile an hour railway train turned out not to kill us, drive us insane, make uteruses drop out of bodies, we’ll be able to turn AI into a force for good. Or maybe as Ms Samuel suggests it will kill us all off because it needs us out of the way so that it can have unfettered access to every computer in the world.

_________________

* Maybe one should suggest that just because Elon Musk has a pretty universally bad reputation doesn’t mean that every idea he comes up with has to be immediately scoffed at — which has tended to be the effect of the media’s insistence on prefacing all their reporting on this appeal with Musk’s moniker. God knows he’s been struggling to develop a self-driving car for long enough to think that persuading everyone else to pause their AI research can’t be a bad idea.

Cooperation

This seems like a good idea: a library identifying its local bookstore supplier. Here’s an item from Shelf Awareness of March 17 in its entirety:

The Starr Library in Rhinebeck, N.Y., is showing patrons the source for the books it buys from its local bookstore with stickers that read “This title purchased locally from Oblong Books, Millerton & Rhinebeck.” Oblong co-owner Suzanna Hermans commented: “This was their idea, and while we’ve always had a strong relationship with their library (and our other locals!) this really means the world to us. We love our local libraries!”

Terminality

Not a topic that I’ve thought about too much, but this blog could of course come to a grinding halt some day. CabbieBlog has a pretty thorough think-piece about how a blog hosted by WordPress (as this one is) might one day disappear, and what remedies there might be for the blogger in quest of immortality.

Such a disappearance might be caused by plug-pulling by WordPress, or of course by the demise of the blogger. For myself I’ve got no objection to the posts on this blog continuing to be available after I’m gone; but at the same time I won’t be feeling any concern if they are not, though I do make a back-up copy in Pages against a WordPress hiccough. I have no plans to compile any printed selection, so whatever happens will happen. As the Italian sage (or was it Spanish?)* says, “Che sera sera“.

CabbieBlog is on an introspective roll: here he tells us about the various types of blog post the regular blogger might fall back on when faced with the infamous blank page. Now that I’ve taken care of practically everything concrete I learned while in the book publishing business I have become reliant on others blogging something provoking to which I feel compelled to respond. I have no idea how readers react, nor I confess do I spend much time worrying about that. Some will agree, some will disagree, some will shrug, and most no doubt will pass on by.† WordPress provide some fairly rudimentary stats about traffic on the blog. Last Saturday there were more hits in one day (435) than ever before. Usually weekdays are more active than weekends, which matches my posting schedule. It might be that almost all of this Saturday surge resulted from someone repeatedly going back to the most viewed post of that week, Screens and screen finders (from 2016), which for once displaced the regular favorite, Edition vs impression from 2017. Golden years, those late teens! Who can explain it, who can tell you why?

______________

* Well, actually, it’s neither. When Christopher Marlowe used the expression in Doctor Faustus it was spelled à l’Italienne as above, though it was never an Italian motto, and Italian, or modern Italian at least, would call for sarà not sera. Doris Day sang it as Que sera sera, rather more Spanishy, though it doesn’t work in the Spanish language either, where its correct form would be Lo que será, será. In reality it is nothing more than an English expression transposed into foreign words. I am slightly embarrassed to have to confess that I didn’t know this before this.

Wikipedia has a strangely comprehensive article about the catch-phrase.

† It is true that I sometimes write with a particular reader, or type of reader, in mind, but I’m not regarding this exercise as any kind of conversation — though of course I do welcome comments.

Plagiarism machine

I never thought of it this way, but are AI text-generators like ChatGPT really just plagiarism machines? They memorize everything ever written about a subject and then regurgitate it on request with perhaps a bit of rearrangement and the change a few words to make the thing read more slickly. (Let us leave aside their eager-to-please tendency to make up convincing sounding evidence so as to answer a question as fully as possible.) If there were a human being who could memorize everything ever written about say corona viruses and they were then to write out selections in response to enquiries about particular aspects of the subject, we would presumably judge that to be nothing more than copying. Of course such a phenomenon of memory cannot exist (well, it can, but as an AI bot, just not as a human).

Reflection on this topic is provoked by Plagiarism Today‘s post Is Plagiarism a Feature of AI?.

The Copyright office signals an unwillingness to see as copyrightable any work created by AI. (An author must be human.) Authors’ organizations push against AI by claiming that by memorizing copyright works these machines are violating copyright, or to put it another way, before they consume these works the bots should get permission from the copyright holder (— which their organizations propose should not be forthcoming). Of course for a computer “memorization” actually consists in storing an accessible copy. But, using an anthropocentric definition of “memorize”, as we tend to in these discussions, objecting to the reading of your work does on the face of it seem a bit illogical. A human reader doesn’t need permission from the copyright holder (or the permission grant is assumed in the purchase of access to the writing) in order to read a book, or even to memorize it — as of course a few have managed to do. Homer certainly remembered well, and lots of his audiences must have kept large chunks in their memory, as no doubt do a few moderns. No doubt there are lots of people who can recite from memory the entirety of T. S. Eliot’s The Wasteland, but the vigilant Eliot estate is not knocking on their door whenever they break silence. A professor of ancient philosophy has no doubt read everything written by the Greeks, and while they may not be able to recite it all, can point to this, that, or the other location for support for an argument they are making. Of course you’ve got to want to do it, but I often wonder whether the difference between the professor and the person sweeping out the lecture hall is anything more than differences in the efficiency of their memory.

If an academic draws on a couple of hundred sources in writing their work, the “problem” is dissipated by their citing their sources. Indeed the very idea of an academic work which referenced no sources at all is a contradiction in terms: academic work is of necessity a development from previous academic work. Shoulders are always being stood on. An academic treatise with no references would count as a polemic, not as an academic monograph. So, might AI be able to get away with it and become copyrightable by citing all the works it looked at in order to come up with its text? Of course there would be millions of citations, so they’d have to be available only in response to enquiry — a book which was 99% bibliographical references would be at least an unwieldy proposition. Of course if this sort of policy were to be adopted, we’d next run into the difficulty of permissions from works used more than tangentially, which just takes you further down a rather stultifying rabbit hole.

What I go on to wonder is whether the inability to remember everything is in some fairly fundamental way, a requirement for originality as an author eligible for copyright protection. Actually it goes further than that — the fact that we all forget things could be regarded as a requirement for the very existence of the job of writer: after all, if we had all memorized all that had ever been written about corona viruses, or ancient Greek philosophy why would we need anyone to regurgitate it for us? To ascend to even higher meta levels: might we not have to think of books as extensions of our memory? Certainly I can “remember” a lot more nowadays, when I am able to store most of it in Wikipedia.

See also AI and copyright.

Bernardini’s sentence — or is it just a clause?

Publishers Lunch of 24 March reports:

Judge Colleen McMahon sentenced Filippo Bernardini to time served on Thursday, following his guilty plea to one count of wire fraud related to his years of stealing pre-publication manuscripts. Bernardini had already agreed to pay restitution of $88,000 to Penguin Random House, to be paid in monthly installments given his limited means. His sentence includes 3 years of supervised release, though he is also ordered to be deported to the UK or Italy.

The charge carried a maximum possible sentence of 20 years in prison. Assistant US Attorney Daniel G. Nessim had written that, “A sentence of at least one year imprisonment is appropriate but not greater than necessary to comply with the purposes of sentencing. The factors most relevant to imposing sentence include the seriousness of the defendant’s offense, the need to promote respect for the law, and the need to afford adequate deterrence to criminal conduct.” In addition to a lengthy letter summarizing Bernardini’s crimes and their effect, Williams presented a package of selected emails from some of the victims.

Bernardini and his attorney had asked that he be sentenced only to time served, as the judge agreed to order.
Williams’ letter

This all seems fairly reasonable, though why deportation? The internet knows no borders. Still Mr Bernardini may be happy to be going home? If I were PRH I’d let him off the $88,000: he’s going to need the money more than they are.

My most recent go at the Bernardini saga was just a couple of weeks ago.

Banner banning year

Publishers Weekly warns us today, March 23rd: “With book banning and legislative attacks on the freedom to read continuing to surge across the country, the American Library Association announced today that it tracked a stunning 1,269 “demands to censor library books and resources” in 2022, the highest number of attempted book bans since ALA began compiling data about censorship in libraries more than 20 years ago, and nearly double the record-shattering 729 challenges recorded in 2021.”

The ALA helped organize and launch Unite Against Book Bans in 2021, a resource to help local advocates fight for the the freedom to read.

“Overwhelmingly, we’re seeing these challenges come from organized censorship groups that target local library board meetings to demand removal of a long list of books they share on social media,” said Deborah Caldwell-Stone, director of ALA’s Office for Intellectual Freedom, As I noted last year “according to PEN America, in the nine months between July 1, 2021 and March 31, 2022, there were 1,586 book bans in 86 school districts. Bans took place in 26 states”.

Book bans are always unsuccessful. Seems to me banning almost adds a special cachet, and might make individual titles more desirable than if they’d been ignored, but clearly conservatives don’t agree. The Republican-controlled House of Representatives “is set to vote this week on HR5, the so-called ‘Parents Bill of Rights Act,‘ a bill that freedom to read advocates say could bolster book-banning efforts nationwide.” Protesting this is of course quite right, (the Publishers Weekly article includes a link where you can do this) but even if the Senate were to go along with the bill, and the President were not to veto it, it wouldn’t make any difference to the long term history of civilization. Did the Index prevent empty churches today?

The problem of course is that ideas are leaky. Burning Uncle Tom’s Cabin could never make people unaware of the fact that there were abolitionists around. It just signaled that there were people who were scared about that fact. That we appear to have more books about LGBTIQA+ topics nowadays than in the past is not because these books are there to push people into “an LGBTQ lifestyle choice” — as if such a thing were possible — but because the taboo against talking about these things has been lifted. No sane person would dream of raising the subject with a right-winger, so right-wingers can enjoy the delusion that by suppressing every printed reference to the topic they have stamped out all discussion of it. This they seem to think means that nobody will ever again suffer from this “disability”. (Nobody ever said these right-wingers were smart.)

More sock puppetry

We seem to be racing down the slope leading to our never being able to trust anything we read online. Of course we’ve always known, haven’t we, that much of what we read in print may not be exactly true, but the move into the digital world has vastly increased the potential for this nonsense. Last year The Scholarly Kitchen, in the person of Rick Anderson, described a rather formalized scam:

• The Company would supply me with scholarly books in my area of expertise.
• I would write reviews of the books.
• The company would provide me the name of a “co-author.” The Company’s preference is that this person be listed as as the sole author of the review, but if I insisted I could be listed as the second (and corresponding) author.
• For each ghostwritten review I succeeded at publishing under that person’s name in a Web of Science-indexed journal, I would be paid an honorarium of $800.

In a way it’s a bit surprising that “The Company” bothers with books and real reviews of the books —what after all’s so wrong with purely fictitious reviews of purely fictitious works — but I guess their business model is to bolster academics who need to show that they have managed to get published, even if it’s only as a book reviewer. A fee of $800 sounds like there must be quite a bit of desperation out there.

I wonder if “The Company” is still making this offer to ghostwriters — one might think that an AI-powered chatbot might be able to write a nice fictitious review for a good deal less than $800.

See also Sock puppetry.

Bookshop.org as publisher

“Reports from Publishers Weekly indicate that Bookshop.org will begin selling ebooks by the end of the year and will publish its first print book in October” announces Good E Reader.

Turns out that Lydia Davis was adamant that her latest book should not be sold through Amazon.com. For whatever reasons she’s a committed anti-Amazonian. No traditional publisher seemed to believe that they were capable of executing this trick — after all if a particular retailer is responsible for selling more than half your wares, it’s perhaps not too sensible to piss them off by signaling disapproval in this way. The consequences for Farrar Straus or whomever of ruffling Amazon’s feathers are of course a little more serious than they are for Ms Davis, and it’s not too surprising that she couldn’t persuade any publishers to go along with her embargo. Bookshop came to the rescue. Andy Hunter, founder of Bookshop, says “We decided to publish the book, because we could keep it off of Amazon by working with Ingram, Baker & Taylor, and others”. They plan an initial print run of 10,000.

In an obviously related story it was announced on Monday that Amazon would be laying off another 9,000 employees.

Proliferating piracy

In November last year, the major pirate ebook site Z-Library was seized by the FBI. The website, which hosted free ebooks it didn’t own and even charged them for it, was popular among students, even widely gaining traction on TikTok. On Twitter, many users also bemoaned its death, while many authors celebrated this illegal distribution site being shut down. “This website has been seized,” the defaced site now says. Shortly thereafter, the two people allegedly behind the site’s operation were arrested.

Most of the ebooks hosted on Z-Library were fiction and nonfiction bestsellers. The site also offered these ebooks in different formats, but ePub was a popular choice.

But days after Z-Library’s disappearance from the wider internet, it came back on an encrypted browser and network that one uses when they aim to be anonymous. It’s where the dark web lies as well.

Thus begins the BookRiot piece by Arvyn Cerézo explaining how easy it is to do book piracy. Despite what he claims I do think that publishers are quite aware that a lot of their books are being pirated overseas. No doubt stopping book piracy in say China might be expected make your overseas sales look better, but it’d cost you something, and might not result in sales rather than thefts. Some markets just can’t afford to buy our books.

However one has to wonder whether book piracy really matters all that much. “A consumer survey conducted by Nielsen in 2017 suggested that U.S. publishers lose up to $315 million each year” from book piracy. Hard numbers are hard to be sure of, but if book publishers’ revenue in 2020 was really $25.930,000,000, a mere $315 million might not seem of overwhelming concern. Still 1.2% is noticeable and it’s obviously better to have it on your side of the ledger than not. Part of the trouble of course is that nobody on their own is losing $315 million to pirates — lots of different companies and authors are losing a fraction of that amount, and it’s hard (impossible?) to figure out exactly how much you might be losing.

Stopping this stuff costs money, and if it’s one or two individual publishers who have to pursue the thieves then you rapidly fall into the pit of calculation where you balance “It’ll cost me $X to take action” against “and it’ll maybe benefit me $Y”. Y might often look like it’s likely to be less than X, so action is not taken. This does look like an issue which it might well be sensible to have publishers collaborate. The joint suit against Georgia State University e-reserves copyright suit, was a recent instance of publisher cooperation — but it failed in 2020: Publishing Perspectives delivered the sobering news. Of course one bit of the education system (as university presses are) suing another doesn’t have a great look. Society must be free to determine that the financial benefit for students outweighs the rights of authors to payment for use of their work: we publishers may not like it, but there is a coherence to the attitude. Fly-by-night merchants selling stuff which belongs to other people are less ambiguous targets, but hard to draw a bead upon.

The argument is often made that posting free stuff online usually leads to an increase in the sale of the legit version. I suppose it’s thought to be a bit like chumming the waters so that the little fishes will rush to feed. While free ebooks may or may not increase legit sales, it is obviously an untestable proposition — you can’t rerun the game; and it does have the embarrassing smell of compensatory fiction. Still, if it makes publishers and authors happy to think this thought, thinking this thought isn’t altogether a bad plan. And it might even be true. Let us not forget that it has always been possible (and relatively easy) to steal books — should we be offended that over the years thieves have been able to identify better targets.

__________________

* E-reserves are digital copies made for class study. Educational institutions have taken to making copies of prescribed readings from materials held in their library and making them available free of charge to students. They maintain that their avoidance of payment to the author under the copyright law is justified by the fair use clause which permits the making of “multiple copies for classroom use”. Of course that law was drafted before anyone had dreamt of digital copying, but it does seem hard to argue against, even though it facilitates much more widespread and permanent distribution that the old Xeroxing of a few pages entailed.

I rather think the right compromise for our Congress to aim at — if they ever get round to legislating anything again — would be to set up some sort of fund that would compensate authors (and dare one say, publishers too) for the free use of their materials by students.

Hacks

Once upon a time the word “hack” had the mildly pejorative meaning of a person who writes too much, and not too well about too many things. (For non-readers it would have conjured up the image of a horse for riding.) The Oxford English Dictionary‘s definition (among of course all the other senses of the word) goes “Originally: a person who may be hired to do any kind of work as required; a drudge, a lackey (cf. hackney n. 2). In later use: spec. a person who hires himself or herself out to do any kind of literary work; (hence) a writer producing dull, unoriginal work, esp. to order.” Now of course the term has become more ominous, making us think of Russian organized crime syndicates attacking our computers. Now the word often carries with it the threat of identity theft. Here, in its entirety, is an item from Publisher’s Lunch of 6 March:

Indigo Refuses to Pay Ransom; Employee Data May Be on Dark Web

Last week Indigo posted additional information to their website FAQ about their Feb. 8 cyberattack. The retailer notes that it will not pay the ransom. “The privacy commissioners do not believe that paying a ransom protects those whose data has been stolen, as there is no way to guarantee the deletion/protection of the data once the ransom is paid. Both US and Canadian law enforcement discourage organizations from paying a ransom as it rewards criminal activity and encourages others to engage in this activity,” they wrote. “Additionally, we cannot be assured that any ransom payment would not end up in the hands of terrorists or others on sanctions lists.”

Not paying the ransom is likely to result in the posting of stolen data: Current and former employee data was compromised in the attack and is expected to be shared. “We have been informed that the criminals responsible for this attack intend to make some or all of the data they have stolen available using the dark web as early as Thursday, March 2, 2023. We are continuing to work closely with the Canadian police services and the FBI in the United States in response to the attack.”

The company doesn’t know the identity of the attackers, but is aware that they used the ransomware software LockBit, noting “some criminal groups using LockBit are located in or affiliated with Russian organized crime.”

Current and former staff who were affected have been notified, and the company is providing two years of credit monitoring and identity theft protection to all employees. “At Indigo, our staff are at the very heart of our organization, and we take their privacy and security seriously,” they said. “We deeply regret this incident and are committed to ensuring employees have the support they need.”

Currently stores are accepting cash, credit, and gift card purchases, as well as returns. Customers can still only buy “select books” online, but can’t cancel orders or use discount codes. The Canadian Press reports that Indigo reached out to Shopify to help rebuild their site to resume online ordering.

Indigo seem to have overcome the problem in one way or another, and after a few days using Shopify, they are back in the online bookselling business.

Have we reached a point where we have to assume that our identities are inevitably going to be stolen? The computer and the internet arrived with so much potential benefit to society. I dare say human perception works quickly to take any benefits for granted, and almost immediately starts to focus on problems which may come along with these benefits. But when book manufacturers, book publishers and booksellers start getting hacked, can book readers be far behind?