Things are going so well, we ought to figure out a way to ruin it all

Plus the row over whether AI can produce art

A couple of things happened this week. First, the Internet Archive lost its appeal against a decision saying it could not lend scanned electronic versions of books owned by it, or by libraries in a network it belonged to. The appeal judgment is pretty blunt, and doesn’t leave the IA much wiggle room for further recourse. So, goodbye to some sort of six-figure sum from the IA’s coffers.

Second, quite separately, I linked in a comment on James O’Malley’s post about his first year of Substacking (free to read) to a Medium post I wrote about how I put together The Overspill, my weekdaily links list; that included a mention of Delicious. “I’d completely forgotten about Delicious!” remarked Katie Lee.

You too had probably forgotten completely about Delicious, if you ever knew what it was. I happen to remember it because it featured in some of the early articles when I took over editing the Guardian’s Technology section in late 2005. Del.icio.us (as it was actually called; and you thought “Twttr” was bad) was snapped up by Yahoo, which used to be a serious company, at the end of 2005 along with Flickr, because it was obvious then that the future consisted of bookmarking sites and uploading photos to a different, or perhaps the same, site.

What sparked my interest about these two events, which may seem totally disparate, is that the problems that the IA and Delicious suffered both came out of a behaviour that so many organisations fall prey to: the desire to get into something new rather than, as the business saying goes, sticking to their knitting. We’re really good at something, the thinking seems to go: now let’s get into something we’re not good at.

To be clear, I’m not talking about pivoting. Pivoting can be necessary; it can be smart; it can be both. A company called Odeo set up in the early 00s to get into a new thing called podcasting; but its founders decided there might be more money in social networking. So they formed a company called Twttr, later Twitter. The owners of YouTube (of whom more later) set it up as a dating site where you’d upload video of yourself; then they decided to let you upload any old video. That was a $1.65bn bright idea.

And now, the mistakes.

Way back when

The Internet Archive and its wonderful Wayback Machine has been a fantastic resource for years, taking snapshots of a colossal variety of web pages which you can then refer back to. It’s the definitive answer to link rot, besides being a terrific way to have a reference that doesn’t depend on the website not changing.

So.. why was the IA being a lending library? According to the appeal judgment, since 2011 it has been scanning books and then allowing people who have accounts with it to download a DRM’d version of the scan. As the judgment puts it,

“This appeal presents the following question: is it ‘fair use’ for a nonprofit organization to scan copyright-protected print books in their entirety and distribute those digital copies online, in full, for free, subject to a one-to-one owned-to-loaned ratio between its print copies and the digital copies it makes available at any given time, all without authorization from the copyright-holding publishers or authors?”

To the dismay of the IA and various of its friends, both the district court and now the appeals court say: no. Based on the Copyright Act “as well as binding Supreme Court and Second Circuit precedent”. The first of those in particular might be hard to shift.

Why, though, did the IA get into the ebook lending business? To quote SwiftOnSecurity on Twitter:

I mean, right? But instead it’s now on the hook for their own costs, and whatever costs the publishers demand for putting them through both the first case and the appeal. You were great at archiving webpages, people, but books have a much more feral bunch backing them up, plus lots of established law. This hasn’t pleased the copyright minimalists: Mike Masnick at Techdirt describes it as “Second Circuit Says Libraries Disincentivize Authors To Write Books By Lending Them For Free”, which I think is overstating things by quite a few degrees. The IA fell foul of the courts not because it was letting people who walked up to its door borrow its physical books, but because (as the judgment says on p15) the IA

“brings to the marketplace a competing substitute” for library eBook licenses, “usurping a market that properly belongs to the copyright holder.”

Libraries in the US have to negotiate ebook licences, and those aren’t free or unlimited, and the IA is in effect commercial because it takes a slice when people buy books through it. Damn those legal niceties!

Which all goes to show how accurate the SwiftOnSecurity tweet was. The IA had a great model going with the web archiving. The leaders of the IA didn’t need to get into lending ebooks. But they wanted to. And that was a mistake.

What does video taste like? Tasty?

Next we come to Delicious. Yahoo bought it at the end of 2005, and then spat it out in April 2011, selling it to the original founders of YouTube.

So, you’re the YouTube founders, you’ve got hundreds of millions of dollars, and now you own one of the web’s most loved bookmarking services. Even though bookmarking wasn’t quite the flavour of the month or even year, there were successful sites offering it. Instapaper (a Marco Arment brainwave) launched in 2008, lett you save content to a cloud service for later reading. Pinboard, run by Maciej Cieglowski, launched in July 2009, simply offered the same API as Delicious had, thus tempting away Delicious users (such as me) who had become increasingly worried over the previous few years about what was happening to the service under Yahoo’s increasingly uninterested ownership.

Great—so what were the loaded YouTube guys going to do with bookmarking in their new company AVOS? By September 2011, a scarce few months later, the answer was clear:

The re-launch of social link sharing site Delicious, now under the stewardship of YouTube founders Steven Chen and Chad Hurley under their AVOS startup banner, is nothing short of a complete, mind-boggling disaster.

How AVOS took a beloved social sharing site and ruined it from stem to stern, and up to this minute have a complete, angry user PR explosion on their hands, is as enlightening as it is hard to watch.

…The new Delicious is essentially doing away with one of the main things the site was great for: tagging and organizing by tags.

Delicious came up with the interesting but seriously misguided idea of forcing users to now categorize and share by using "stacks" - among many other forced changes.

There was a colossal list of things they were “working on”. And the list of things that didn’t work was long:

The RSS feeds were broken, the password reset was broken, browser extensions are still broken, tag bundles are gone (users put a lot of work into these), search by date is gone and search returns are not chronological, users are now unable to edit their tags...

Among the many, many mistakes AVOS made was not knowing what people wanted to use Delicious for. In taking it over, they failed to make the character “/” searchable—and this turned out to be their downfall, because Delicious had been hugely popular with fandoms who wrote fanfic, and fanfic relies on “/”.

In 2012, AVOS got an undisclosed amount of venture capital funding to help grow their new baby—would it add video somehow? Get into proper social networking? All that money sloshing around! Surely that would make something happen?

But the trouble was that Chad Hurley and Steve Chen had no obvious idea for what they wanted Delicious to be or do. All that experience getting a video startup to work counted for nothing when trying to get an established bookmarking service to continue to work.

And so nothing happened, except Delicious faded away into irrelevance—while Pinboard grew because, as this Verge article from 2017 explains, it actually listened to what the fandoms infuriated by Delicious’s lack of usefulness wanted.

And in 2017, Delicious completed its life cycle: having been started in a room by Joshua Schachter and Peter Gadjokov in 2003, and sold to Yahoo in 2005 for perhaps $10m, and then offloaded by Yahoo to AVOS in 2011 for perhaps $1m, Chen and Hurley admitted defeat in 2017 and sold it to Pinboard for $35,000. Because it’s Cieglowski who has the smart idea: just do something, but do it so well that people stick with you. (Personally, I’ve been using Pinboard since 2010 or so.)

Don’t try this at home. Or work

Those are only two examples, at length, of people trying things they shouldn’t have, and discovering the hard way why not. But this does seem to be one of those axiomatic behaviours. Rather like every app expands until it includes messaging, and every messaging app expands until it includes the behaviours of any number of other apps, and the natural endpoint of unbundling (such as cable TV) is not complete atomisation but re-bundling, so organisations get into things they’re not good at because they aren’t doing them yet. If you need a quick illustration, think of Apple: for years it has made handsome profits on hardware—phones and tablets and music players and headphones—so obviously it needed to get into the video streaming business and commission its own programmes at huge expense and nothing even approaching profit. (Cooler heads are now prevailing. Fine, but don’t take away Slow Horses.)

More proximately, one looks at what used to be Twitter—really good at getting short bursts of text, then pictures, then also short video—to people all round the world, and then you wonder at its current trajectory where it will allegedly become a banking app that will let people send money all over the place. (Do exchange rates come into this? Facebook came very unstuck when it tried to overcome that with its own digital currency called Libra.) Plus also a full streaming app like.. YouTube? Let’s guess how well that is likely to go, given how well the “blue tick” verification scheme was handled.

There must be other examples: I’d love to hear them. Suggest away in the comments.

---

Glimpses of the AI tsunami

(Of the what? Read here. And then the update.)

• Ted Chiang in the New Yorker insists that AI can’t produce art because (I paraphrase) it doesn’t know about making choices.

• Ted Chiang is wrong about what defines the difference between “art” and “not-art” because it’s not about choice but (I paraphrase) domain space.

• NaNoWriMo, the NAtional—actually interNAtional—effort which encourages people to write (or finish) a NOvel in the MOnth of NOvember, has an odd announcement about using AI to write which has displeased almost everyone who reads it, except perhaps AIs.

• LLMs can implant false memories about things you’ve seen through question prompts.

---

• You can buy Social Warming in paperback, hardback or ebook via One World Publications, or order it through your friendly local bookstore. Or listen to me read it on Audible.

You could also sign up for The Overspill, a daily list of links with short extracts and brief commentary on things I find interesting in tech, science, medicine, politics and any other topic that takes my fancy.

• I’m the proposed Class Representative for a lawsuit against Google in the UK on behalf of publishers. If you sold open display ads in the UK after 2014, you might be a member of the class. Read more at Googleadclaim.co.uk. (Or see the press release.)

• Back next week! Or leave a comment here, or in the Substack chat, or Substack Notes, or write it in a letter and put it in a bottle so that The Police write a song about it after it falls through a wormhole and goes back in time.