Werd I/O

This is my monthly roundup of the links and media I found interesting. Do you have suggestions? Let me know!

Apps + Websites

DALL·E 3. Once again, this looks completely like magic. Very high-fidelity images across a bunch of different styles. The implications are enormous.

Photoshop for Web. Insanely good. It blows my mind that this can be done on the web platform now.

Privacy Party. This is really good: a browser extension (for Chrome-based browsers) that goes through your social networks and helps you update your settings to optimize for privacy and security. Really well-executed.

Notion web Clipper - Klippper. I’m a heavy Notion web clipper user, but this is far better for my needs. I was worried I’d need to build it myself. Luckily: no!

Mastodon 4.2. Lots of good new changes here - and in particular a much-needed search overhaul. My private instance is running the latest and I like it a lot.

Notable Articles

AI

How the “Surveillance AI Pipeline” Literally Objectifies Human Beings. “The vast majority of computer vision research leads to technology that surveils human beings, a new preprint study that analyzed more than 20,000 computer vision papers and 11,000 patents spanning three decades has found.”

California governor vetoes bill banning robotrucks without safety drivers. The legislation passed with a heavy majority - this veto is a signal that Newsom favors the AI vendors over teamster concerns. Teamsters, on the other hand, claim the tech is unsafe and that jobs will be lost.

ChatGPT Caught Giving Horrible Advice to Cancer Patients. LLMs are a magic trick; interesting and useful for superficial tasks, but very much not up to, for example, replacing a trained medical professional. The idea that someone would think it’s okay to let one give medical advice is horrifying.

AI data training companies like Scale AI are hiring poets. These poets are being hired to eliminate the possibility of being paid for their own work. But I am kind of tickled by the idea that OpenAI is scraping fan-fiction forums. Not because it’s bad work, but imagine the consequences.

John Grisham, other top US authors sue OpenAI over copyrights. It will be fascinating to see the outcome of this - which, in turn, will set a precedent for how commercial data can be used to train AI (and other software systems) going forward.

Who blocks OpenAI? “The 392 news organizations listed below have instructed OpenAI’s GPTBot to not scan their sites, according to a continual survey of 1,119 online publishers conducted by the homepages.news archive. That amounts to 35.0% of the total.”

Microsoft announces new Copilot Copyright Commitment for customers. “As customers ask whether they can use Microsoft’s Copilot services and the output they generate without worrying about copyright claims, we are providing a straightforward answer: yes, you can, and if you are challenged on copyright grounds, we will assume responsibility for the potential legal risks involved.”

Our Self-Driving Cars Will Save Countless Lives, But They Will Kill Some of You First. “In a way, the people our cars mow down are doing just as much as our highly paid programmers and engineers to create the utopian, safe streets of tomorrow. Each person who falls under our front bumper teaches us something valuable about how humans act in the real world.”

Climate

EVs are a climate solution with a pollution problem: Tire particles. Another reason why the really sustainable solution to pollution from cars is better mass transit.

Revealed: top carbon offset projects may not cut planet-heating emissions. “The vast majority of the environmental projects most frequently used to offset greenhouse gas emissions appear to have fundamental failings suggesting they cannot be relied upon to cut planet-heating emissions, according to a new analysis.”

Earth ‘well outside safe operating space for humanity’, scientists find. “This update finds that six of the nine boundaries are transgressed, suggesting that Earth is now well outside of the safe operating space for humanity.” No biggie.

Why the United States undercounts climate-driven deaths. Another way the effect of the climate crisis is understated: climate deaths are undercounted. Changing this state of affairs is possible but requires effort, training, and resources. In the meantime, many people still don’t understand how serious the crisis actually is.

Culture

Nature TTL Photographer of the Year 2023: Winners Gallery. Every image here is stunning.

‘The scripts were the funniest things I’d ever read’: the stars of Peep Show look back, 20 years later.Before there was Succession, there was Peep Show. A brilliant piece of TV that launched a bunch of careers. If you haven’t seen it, give yourself the gift of checking it out.

The Berkeley Hotel hostage. I know people who worked with Douglas Adams and I’m incredibly envious of them. He seems like someone I would have really enjoyed meeting - and his books (all of them) were a huge part of my developing psyche. This story seems so human, so relatable. Trapped by his success, in a way.

Refusing to Censor Myself. A less-discussed problem with book bans: publishers will self-censor, as they did here by requiring the removal of the word “racism” in the context of internment camps.

Writer Sarah Rose Etter on not making things harder than they need to be. I found this interview fascinating: definitely a writer I look up to, whose work I both enjoy and find intimidatingly raw. And who happens to have a very similar day job to me.

Democracy

FTC Sues Amazon for Illegally Maintaining Monopoly Power. “Amazon’s ongoing pattern of illegal conduct blocks competition, allowing it to wield monopoly power to inflate prices, degrade quality, and stifle innovation for consumers and businesses.” Whatever happens here, it will be meaningful. It’s also nice to see the FTC actually wielding its antitrust powers.

Intuit Pushing Claim That Free Tax-Filing Program Would Harm Black Taxpayers. Intuit has a stranglehold on how taxes are filed in America. For what? Many other countries just have an easy to use tax portal of their own. This is a business that shouldn’t even need to exist.

Migrants tracked with GPS tags say UK feels like ‘an outside prison’. I had no idea Britain was fitting migrants and asylum seekers with ankle bracelets and surveilling them to this level. It seems impossible that this is something people would think is right and just. The dystopian cruelty is mind-boggling.

An endless battle for the rights of the underclass. Every word of this, but particularly: “Cultural warfare was a political ploy designed to keep workers from recognizing our common ground and banding together against corporate abuses and thefts.”

US economy going strong under Biden – Americans don’t believe it. It’s how we measure the economy, stupid.

What Mitt Romney Saw in the Senate. A fascinating read that makes me want to check out the full book, which seems to me like an attempt by Romney to save the Republican Party from Trumpism (as well as, let’s be clear, his own reputation). Wild anecdote after wild anecdote that highlights the cynicism of Washington political life.

Never Remember. The best thing I read on the anniversary of 9/11 by far. It feels cathartic to read. But it’s also so, so sad.

New Elon Musk biography offers fresh details about the billionaire's Ukraine dilemma. If I was building technology to let people watch Netflix and check their email from remote locations, I would also be upset about it being used for drone strikes. But if that’s the case, you shouldn’t be deploying your tech to the military in the first place. Nor should you be making strategic military decisions of your own.

Majority of likely Democratic voters say party should ditch Biden, poll shows. No surprises here. We need more progressive change than we’re getting. But obviously, if it’s Biden v Trump, there’s only one choice.

AOC urges US to apologize for meddling in Latin America: ‘We’re here to reset relationships’. Yes. Absolutely this. And everywhere.

Health

The Anti-Vax Movement Isn’t Going Away. We Must Adapt to It. Depressing. I agree that vaccine denial is not going away, and that we need to find other ways to mitigate outbreaks. But what a sad situation to be in.

Labor

Remote work may help decrease sexual assault and harassment, poll finds. “About 5 percent of women who were working remotely reported instances in that time, compared with 12 percent of in-person women workers. Overall, only 5 percent of remote workers reported instances in the past three years, compared with 9 percent of those who work fully or mostly in person.”

Working mothers reach record high, above pre-pandemic levels. Flexible work from home policies have allowed more mothers with young children to join the workforce than ever before. Yet another reason why these policies are positive for everyone and should not just stick around but be significantly expanded.

Media

Amanda Zamora is stepping down as publisher at The 19th. Amanda is absolutely fearless and I was privileged to work with her. As co-founder of The 19th, she was an absolutely core part of what it became: both a strategist and culture instigator. What she does next will certainly change media; I’ll be cheerleading.

Failing Without Knowing Why: The Tragedy Of Performative Content. Thought-provoking for me: particularly as someone who thinks through ideas through writing. But perhaps that writing doesn’t need to be in public, in front of an audience.

How I approach crafting a blog post. “I don’t think I’ve seen someone walk through their process for writing a blog post, though.” I love this breakdown! Tracy’s structured process shows up in the quality of her posts. I love the thoughtfulness here.

In defense of aggressive small-town newspapers. This: “The prevalence of “news deserts” has apparently led some to think it’s normal for neighborhood news outlets to function as lapdogs rather than watchdogs.” The purpose of journalism is to investigate in the public interest.

In the AI Age, The New York Times Wants Reporters to Tell Readers Who They Are. I think this is the right impulse: people tend to follow and trust individual journalists, not publications. Building out profiles and establishing more personal relationships helps build that trust.

Counting Ghosts. “Web analytics sits in the awkward space between empirical analysis and relationship building, failing at both, distracting from the real job to be done: making connections, in whatever form that means for our project.”

Publisher wants $2,500 to allow academics to post their own manuscript to their own repository – Walled Culture. The open access movement is an important way academics can fight back against predatory publishers for the good of human knowledge everywhere - but the publishers are still out there, grifting.

A New Low: Just 46% Of U.S. Households Subscribe To Traditional Cable TV. I’ve lived in the US for twelve years, and at no point have I even been tempted by traditional cable. Every time I encounter it, I wonder why people want it. It’s a substandard, obsolete product. So this is no surprise.

The Ad Industry Bailed On News. Can An AI Solution Offer A Way Back? Services like this become single points of failure with outsize power over the journalism industry. It’s a bad idea. No one entity should be the arbiter of bias in news or where a buyer should put their money. For one thing, who watches that entity’s own inevitable bias? And if you’re offering AI as a bias-free solution, you’ve already lost.

Zine: How We Illustrate Tech (and AI) at The Markup. Lovely!

White House to send letter to news execs urging outlets to 'ramp up' scrutiny of GOP's Biden impeachment inquiry 'based on lies'. I couldn’t be less of a fan of the current Republican Party but I hate this. The White House should not be sending letters to the media encouraging them to do anything. That’s not the sort of relationship we need our journalistic media to have.

Snoop Dogg can narrate your news articles. Snoop Dogg gimmick aside, this is actually pretty neat, and useful. I’d also like the opposite: sometimes I want to read podcasts. Different contexts demand different media; I wish content itself could be more adaptable.

Non-news sites expose people to more political content than news sites. Why? Two thirds of the political content people consume come from non-news sites. And most of the news content people read is not overtly political. Instead, it’s mostly coming from entertainment - which has no ethical need to report factually.

Naomi Klein's "Doppelganger". “Fundamentally: Klein is a leftist, Wolf was a liberal. The classic leftist distinction goes: leftists want to abolish a system where 150 white men run the world; liberals want to replace half of those 150 with women, queers and people of color.”

Society

US surgeons are killing themselves at an alarming rate. One decided to speak out. “Somewhere between 300 to 400 physicians a year in the US take their own lives, the equivalent of one medical school graduating class annually.”

Oxford University is the world’s top university for a record eighth year. This presumably means that the Turf Tavern is the best student pub in the world.

Britain’s attitude to refugees shows, once again, that it’s a colonial nation. “Hostile immigration policy stokes racism but the foundation it builds upon itself is racist and maintains a ‘colonial present’. Through dealing with migrants like pests, who deserve to be locked away in a prison barge, the British government continues to ignore the fact that, “Borders maintain hoarded concentrations of wealth accrued from colonial domination.”″

19th News/SurveyMonkey poll: The State of Our Nation. Lots of interesting insights in this poll, including on nationwide attitudes to gender-affirming care (only 29% of Republicans think their party should focus on it) and gun control (82% of Americans want to restrict access in domestic abuse cases).

Victims of forced sterilization in California prisons entitled to reparations. One thing I learned from this story is that forced sterilization of inmates has still been widespread in the 21st century in America. Ghoulish.

Unconditional cash transfers reduce homelessness. It turns out that if you give homeless people money as assistance, it really helps them. This is something society should do.

Startups

Why Starting Your Investor Updates With “Cash on Hand” Information is a Major Red Flag Right Now. It’s Maybe the Only Thing Worse Than Not Sending Updates at All. I appreciated this succinct discussion on using venture dollars well from Hunter Walk. In particular, this: “Startups spend a $1 to ultimately try and create more than $1 of company. If you do that repeatedly and efficiently we will all make money together.” Too many founders still think of investment as being akin to a grant.

Technology

Meta in Myanmar, Part I: The Setup. “By that point, Meta had been receiving detailed and increasingly desperate warnings about Facebook’s role as an accelerant of genocidal propaganda in Myanmar for six years.” We need more discussion of this - I’m grateful for this four-part series.

Optimizing for Taste. A solid argument against A/B testing. A lot of it comes down to this: “It fosters a culture of decision making without having an opinion, without having to put a stake in the ground. It fosters a culture where making a quick buck trumps a great product experience.” I agree.

Meredith Whittaker reaffirms that Signal would leave UK if forced by privacy bill. Signal on UK privacy law: “We would leave the U.K. or any jurisdiction if it came down to the choice between backdooring our encryption and betraying the people who count on us for privacy, or leaving.” Good.

U.S. Counterintel Buys Access to the Backbone of the Internet to Hunt Foreign Hackers. “The news is yet another example of a government agency turning to the private sector for novel datasets that the public is likely unaware are being collected and then sold.”

Digital Disruption: Measuring the Social and Economic Costs of Internet Shutdowns & Throttling of Access to Twitter. This report found that removing access to Twitter created significant economic and social impacts. Question: are some of these now replicated with the switch to X?

Build Great Software By Repeatedly Encountering It. This is really important, and why we talk about “eating your own dogfood”. If you don’t use what you build, you can’t build anything great.

EV charging infrastruture is a joke – Brad Barrish. Non-Tesla EV charging infrastructure is awful. It’s good that Tesla has opened the standard, but it’s not good that the only really viable charging infrastructure is owned by one company. It needs to be fixed.

The #ViewSource Affordance. I strongly agree with this. “View source” has been an important part of the culture of the web since the beginning. Obfuscating that source or removing the option does damage to its underlying principles and makes the web a worse place. I like the comparison to the enclosure movement, which seems apt.

Online Safety Bill: Crackdown on harmful social media content agreed. This is a horrendous bill that is designed to encourage self-censorship, including around topics like “illegal immigration”, as well as vastly deepen surveillance on internet users. And Britain passing it will likely embolden other nations to try the same.

WordPress blogs can now be followed in the fediverse, including Mastodon. I’d prefer if this was default WordPress functionality - but the big lede is buried here. Hosted WordPress sites are getting fediverse compatibility. That’s a huge deal.

Finishing With Twitter/X. Who at the intersection of tech and politics is still posting on Twitter? And should they be? A good breakdown.

Unity has changed its pricing model, and game developers are pissed off. As with API pricing changes across social media, these tiers disproportionately penalize indie developers. The message is clear: they don’t want or need those customers. In a tighter economy, much of technology is re-organizing around serving bigger, wealthier players.

Silicon Valley's Slaughterhouse. “Andreessen wasn’t advocating for a tech industry that accelerates the development of the human race, or elevates the human condition. He wanted to (and succeeded in creating) a Silicon Valley that builds technology that can, and I quote, “eat markets far larger than the technology industry has historically been able to pursue.””

Google vet wants to turn your hot water heater into a "virtual power plant". I really need this for my home, and I suspect my entire region needs it. This could do a lot of good and be the start of something much bigger using virtual power plants as a platform.

It’s Official: Cars Are the Worst Product Category We Have Ever Reviewed for Privacy. Every modern car brand abuses your personal information. 84% sell your data (including where you go and when). 56% will share it with law enforcement without a warrant. And none of them have demonstrably adequate security.

Tucson's Molly Holzschlag, known as 'the fairy godmother of the web,' dead at 60. Rest in peace, Molly. We’ve lost one of the really good people who made the web better.

I learned to read and write on computers.

Our first home computer, the Sinclair ZX81, had BASIC shortcuts built into the keyboards: you could hit a key combination and words like RUN, THEN, and ELSE would spit out onto the screen. I wrote a lot of early stories using those building blocks.

Our second, the Atari 130XE, had similar BASIC instructions, but also had a much stronger software ecosystem. In one, you would type a rudimentary story, and 8-bit stick figure characters would act it out on screen. “The man walks to the woman”; “The wumpus eats the man.”

We never had a games console in the house, much to my chagrin, although the Atari could take games cartridges, and I once got so far in Joust that the score wrapped back around to 0. But mostly, I used our computers to write stories and play around a little bit with simple computer programming (my mother taught me a little BASIC when I was five).

We walk our son to daycare via the local elementary school. This morning, as we wheeled his empty stroller back past the building, a school bus pulled up outside and a stream of eight-year-olds came tumbling out in front of us. As we stood there and watched them walk one by one into the building, I saw iPhone after iPhone after iPhone clutched in chubby little hands. Instagram; YouTube; texting.

It’s obvious that he’ll get into computers early: he’s the son of someone who learned to write code at the same time as writing English and a cognitive scientist who does research for a big FAANG company. Give him half a chance and he’ll already grab someone’s phone or laptop and find modes none of us knew existed — and he’s barely a year old. The only question is how he’ll get into computers.

I’m adamant with him, as my parents were with me, that he should see a computer as a creation, not a consumption device. At their best, computers are tools that allow children to create things themselves, and learn about the world in the process. At their worst, they’re little more than televisions, albeit with a near-infinite number of channels, that needlessly limit your horizons. For many kids, social media is such a huge part of of their life that being an influencer is their most hoped-for job. No thank you: not for my kid.

But, of course, if we can steer away from streaming media and Instagram’s hollow expectations, there’s a ton of fun to be had. This is one area where I think generative AI could be genuinely joyful: the fun that I had writing stories for those 8-bit stick figures, transposed to a whole universe of visual possibilities. That is, of course, unless using those tools prevents him from learning to draw himself.

He’s entering a very different cultural landscape where computers occupy a very different space. Those early 8-bit machines were, by necessity, all about creation: you often had to type in a BASIC script before you could use any software at all. In contrast, today’s devices are optimized to keep you consuming, and to capture your engagement at all costs. Those iPhones those kids were holding are designed to be addiction machines.

Correspondingly, our role as parents is to teach responsible use. If we are to be good teachers, that also means we have to demonstrate responsible use: something I am notoriously bad at with my own phone. I’ve got every social network installed. I sometimes lose time to TikTok. I’m a slave to my tiny hand-computer in every way I possibly can be. I tell myself that I need to know how it all works because of what I do for a living, but the real truth is, I love it. I don’t need to be on social media; I don’t need to be a part of the iPhone Upgrade Program. I just am.

I think responsible use means dialing up the ratio of creation to consumption for me, too. If I’m to convey that it’s better to be an active part of shaping the world than just being a passive consumer of it, that’s what I have to do. This is true in all things — a core, important lesson is that there isn’t one way to do things, and life is richer if you don’t follow the life templates that are set out for us — but in some ways I feel it most acutely in our relationship to technology.

There will certainly be peer pressure. His friends will have iPhones. I don’t think withholding technology is the right thing to do: consider those kids whose parents never let them have junk food, who then go out and have as much junk food as possible as soon as they can. Instead, if he has an iPhone, he will learn how to make simple iPhone apps. You’d better believe that he’ll learn how to make websites early on (what kind of indieweb advocate would I be otherwise?). He will be writing stories and editing videos and making music. And, sure, he’ll be consuming as part of that — but, in part, as a way to get inspired about making his own things.

These days, creating also means participating in online conversations. As he gets older, we’ll need to have careful discussions about the ideas he encounters. I’m already imagining that first conversation about why Black Lives Matter is an important movement and how to think about right-wing content that seeks to minimize other people. I don’t want our kid to be a lurker who thinks people should be happy with what they get; I want him to feel like the world is his oyster, and that he can help change it for the better. Our devices can be a gateway to bigger ideas, or they can be a path to a constrained walled garden of parochial thought. It all requires guidance and trust.

The computer revolution happened between my birth and his. Realizing so makes me feel as old as dust, but more importantly, it opens up a new set of parental responsibilities. I want to help him be someone who creates and affects the world, not someone who lets the world happen to him. And there’s so much world to see.

It’s no surprise to anyone that I prefer reading peoples’ long-form thoughts to tweets or pithy social media posts. Microblogging is interesting for quick, in-the-now status updates, but I find myself craving more nuance and depth.

Luckily, Blogging is enjoying a resurgence off the back of movements like the Indieweb (at one end of the spectrum) and platforms like Substack (at the other), and far more people are writing in public on their own sites than they were ten years ago. Hooray! This is great for me, but how do I find all those sites to read?

I figured that the people I’m connected to on Mastodon would probably be the most likely to be writing on their own sites, so I wondered if it was possible to subscribe to all the blogs of the people I followed.

I had a few criteria:

1. I only wanted to subscribe to blogs. (No feeds of updates from GitHub, for example, or posts in forums.)
2. I didn’t want to have to authenticate with the Mastodon API to get this done. This felt like a job for a scraper — and Mastodon’s API is designed in such a way that you need to make several API calls to figure out each user’s profile links, which I didn’t want to do.
3. I wanted to write it in an hour or two on Sunday morning. This wasn’t going to be a sophisticated project. I was going to take my son to the children’s museum in the afternoon, which was a far more important task.

On Mastodon, people can list a small number of external links as part of their profile, with any label they choose. Some people are kind enough to use the label blog, which is fairly determinative, but lots don’t. So I decided that I wanted to take a look at every link people I follow on Mastodon added to their profiles, figure out if it’s a blog I can subscribe to or not, and then add the reasonably-bloggy sites to an OPML file that I could then add to an RSS reader.

Here’s the very quick-and-dirty command line tool I wrote yesterday.

Mastodon helpfully produces a CSV file that lists all the accounts you follow. I decided to use that as an index rather than crawling my instance.

Then it converts those account usernames to URLs and downloads the HTML for each profile. While Mastodon has latterly started using JavaScript to render its UI — which means the actual profile links aren’t there in the HTML to parse — it turns out that it includes profile links as rel=“me” metatags in the page header, so my script finds end extracts those using the indieweb link-rel parser to create the list of websites to crawl.

Once it has the list of websites, it excludes any that don’t look like they’re probably blogs, using some imperfect-but-probably-good-enough heuristics that include:

1. Known silo URLs (Facebook, Soundcloud, etc) are excluded.
2. If the URL contains /article, /product, and so on, it’s probably a link to an individual page rather than a blog.
3. Long links are probably articles or resources, not blogs.
4. Pages with long URL query strings are probably search results, not blogs.
5. Links to other Mastodon profiles (or Pixelfed, Firefish, and so on) disappear.

The script goes through the remaining list and attempts to find the feed for each page. If it doesn’t find a feed I can subscribe to, it just moves on. Any feeds that look like feeds of comments are discarded. Then, because the first feed listed is usually the best one, the script chooses the first remaining feed in the list for the page.

Once it’s gone through every website, it spits out a CSV and an OPML file.

After a few runs, I pushed the OPML file into Newsblur, my feed reader of choice. It was able to subscribe to a little over a thousand new feeds. Given that I’d written the script in a little over an hour and that it was using some questionable tactics, I wasn’t sure how high-quality the sites would be, so I organized them all into a new “Mastodon follows” folder that I could unsubscribe to quickly if I needed to.

But actually, it was pretty great! A few erroneous feeds did make it through: a few regional newspapers (I follow a lot of journalists), some updates to self-hosted Git repositories, and some Lemmy feeds. I learned quickly that I don’t care for most Tumblr content — which is usually reposted images — and I found myself wishing I’d excluded it. Finally, I removed some non-English feeds that I simply couldn’t read (although I wish my feed reader had an auto-translate function so that I could).

The upshot is that I’ve got a lot more blogs to read from people I’ve already expressed interest in. Is the script anything close to perfect? Absolutely not. It it shippable? Not really. But it did what I needed it to, and I’m perfectly happy.

Tracy Durnell celebrates 20 years of blogging:

A blog is a much nicer place to publish than social media, sparking fewer but more meaningful interactions. Blogging allows writers a more forgiving pace with slower conversation. On their blog, people can be themselves instead of playing to an audience and feeling judged — a place to escape the pressures of one-upmanship and signaling, the noise of the ever-demanding attention economy, and the stress of hustle culture.

It’s a huge achievement, to be sure, and I couldn’t agree more with Tracy’s sentiment here. Congratulations, Tracy!

I’m a little jealous that she can pinpoint an anniversary date. For me, it depends on how you judge: I had a hand-rolled blog of sorts when I went to university in 1998, but was it really a blog? I definitely had a public Livejournal in 2001, but was that a blog? How about blog I used to keep on Elgg dot net (now a domain squatter, may it rest in peace)? My old domain, benwerd.com, dates back to 2006, and my current one, werd.io, only goes back to 2013. It’s a bit of a messy history, with stops and false starts.

On the other hand, I know people who have posted to the same domain for almost as long as they’ve been online. I don’t know if I can match that sort of dedication - or a commitment to even having a continuous identity for all that time. Am I the same person I was 20+ years ago? A little bit yes, but mostly not really. The idea of joining up my life online on a long-term basis is actually quite daunting.

Tracy links to Mandy Brown’s piece on writers vs talkers, which also deeply resonates: I’m a writer. I hate being drawn into making decisions in ad hoc meetings. I want to write my thoughts down, structure them, and then come to a conclusion after getting feedback and iterating. Perhaps that’s why blogging early appeals to me so much: I can put out ideas and very quickly engage in conversations about them that pushes my thinking along.

Blogging might seem like a solitary activity, but it’s very, very social. Even the name — a pun derived from weblog = we blog — is about community. Writing for 20 years also means building community for that long.

Here’s to the next 20!

The way my link blog works is like this:

I save an article, website, or book I thought was interesting to a database in Notion using the web clipper, together with a description and a high-level category. (These are Technology, Society, Democracy, and so on.) I also have a checkbox that designates whether the link is something I’d consider business-friendly.

Zapier watches for new links. When it finds one, it publishes it to my website using the micropub protocol. (My website then tries to send a webmention to that site to let it know I’ve linked to them.)

Then, it publishes the link to my Mastodon profile using the top-level category as a hashtag. If the link is to a book, it also adds the bookstodon hashtag.

Following that, it publishes to all my other social networks via Buffer, without the hashtag. (The exception is my Bluesky profile, which I had to write some custom API code for). If the business-friendly box was checked, that includes publishing to my LinkedIn profile.

If I’m feeling particularly motivated, I’ll copy and paste the link to my Threads profile, but because there’s no API, it’s a fully manual process. Which means I usually don’t.

Very occasionally, Zapier will pick up a link before the Notion entry has fully saved, which means that links post without a description or a category. Then I either shrug my shoulders and accept that I have some weird posts on my timeline, or I go back and edit or repost each and every one.

Because of this bug, I’ve thought about writing my own code to do all of the above on my server. It would work the exact way I want it to be. It would be cheaper, too: I pay for Zapier every month, and the cost adds up.

But while I could do this, and the up-front cost would certainly be lower, what if something goes wrong? Let’s say LinkedIn changes the way their API works. If I wrote the connection myself, I would need to keep my code up to date every time this happened — and, in turn, stay on top of codebase changes for every single social media platform I used.

And the truth is: I’m tired, friends. I want to be really careful about the amount of code I set myself up to maintain. It might seem like a simple script now, but over time I build up more and more simple scripts and, cumulatively, I end up buried in code.

As I get older, I find myself optimizing that cost more and more. I’d much rather pay something up-front that saves me a ton of time and cognitive overhead, because both of these things are at such an enormous premium for me.

I could also just not post to those social media accounts, or do it fully-manually, but there’s something really satisfying about publishing once and syndicating everywhere I’m connected to people. I could save my links straight to something like Buffer, but I also like having my categorized database of everything I’ve shared. And Notion makes it easy to save links across my devices (I’m sometimes on my phone, sometimes on my laptop, sometimes on my desktop).

So I’m keeping Zapier, at least for now. I like keeping my links, and I like sharing them. And, more than anything else, I like not having to maintain the code that does it.

Claire Anderson (hi Claire!) asked me to break down the economics of AI. How is it going to make money, and for whom?

In this post I’m not going to talk too much about how the technology works, and the claims of its vendors vs the actual limitations of the products. Baljur Bjarnason has written extensively about that, while Simon Willison writes about building tools with AI and I recommend both of their posts.

The important thing is that when we talk about AI today, we are mostly talking about generative AI. These are products that are capable of generating content: this could be text (for example, ChatGPT), images (eg Midjourney), music, video, and so on.

Usually they do so in response to a simple text prompt. For example, in response to the prompt ‌Write a short limerick about Ben Werdmuller asking ChatGPT to write a short limerick about Ben Werdmuller, ChatGPT instantly produced:

Ben Werdmuller pondered with glee,
“What would ChatGPT write about me?”
So he posed the request,
In a jest quite obsessed,
And chuckled at layers, level three!

Honestly, it’s pretty clever.

While a limerick isn’t particularly economically useful, you can ask these technologies to write code for you, find hidden patterns in data, highlight potential mistakes in boilerplate legal documents, and so on. (I’m personally aware of companies using it to do each of these things.)

Each of these AI products is powered by a large foundation model: deep learning neural networks that are trained on vast amounts of data. In essence, the neural network is a piece of software that ingests a huge amount of source material and finds patterns in it. Based on those patterns and the sheer amount of data involved, it can statistically decide what the outcome of a prompt should be. Each word of the limerick above is what the model has decided is the most probably likely next piece of the output in response to my prompt.

The models are what have been called stochastic parrots: their output is entirely probabilistic. This kind of AI isn’t intelligence and these models have no understanding of what they’re saying. It’s a bit like a magic trick that’s really only possible because of the sheer amount of data that’s wrapped up in the training set.

And here’s the rub: the training set is a not insignificant percentage of everything that’s ever been published by a human. A huge portion of the web is there; it’s also been shown that entire libraries of pirated books have been involved. No royalties or license agreements have been paid for this content. The vast majority of it seems to have been simply scraped. Scraping publicly accessible content is not illegal(and nor should it be); incorporating pirated books and licensed media clearly is.

Clearly if you’re sucking up everything people have published, you’re also sucking up the prejudices and systemic biases that are a part of modern life. Some vendors, like OpenAI, claim to be trying to reduce those biases in their training sets. Others, like Elon Musk’s X.AI, claim that reducing those biases is tantamount to training your model to lie. He claims to be building an “anti-woke” model in response to OpenAI’s “politically correct” bias mitigation, which is pretty on-brand for Musk.

In other words, vendors are competing on the quality, characteristics, and sometimes ideological slant of their models. They’re often closed-source, giving the vendor control over how the model is generated, tweaked, and used.

These models all require a lot of computing power both to be trained and to produce their output. It’s difficult to provide a service that offers generative AI to large numbers of people due to this need: it’s expensive and it draws a lot of power (and correspondingly has a large environmental footprint).

Between the closed nature of the models, and the computing power required to run them, it’s not easy to get started in AI without paying an existing vendor. If a tech company wants to add AI to a product, or if a new startup wants to offer an AI-powered product, it’s much more cost effective to piggyback on another vendor’s existing model than to develop or host one of their own. Even Microsoft decided to invest billions of dollars into OpenAI and build a tight partnership with the company rather than build its own capability.

The models learn from their users, so as more people have conversations with ChatGPT, for example, the model gets better and better. These are commonly called network effects: the more people that use the products, the better they get. The result is that they have even more of a moat between themselves and any competitors over time. This is also true if a product just uses a model behind the scenes. So if OpenAI’s technology is built into Microsoft Office — and it is! — its models get better every time someone uses them while they write a document or edit a spreadsheet. Each of those uses sends data straight back to OpenAI’s servers and is paid for through Microsoft’s partnership.

What’s been created is an odd situation where the models are trained on content we’ve all published, and improved with our questions and new content, and then it’s all wrapped up to us as a product and sold back to us. There’s certainly some proprietary invention and value in the training methodology and APIs that make it all work, but the underlying data being learned from belongs to us, not them. It wouldn’t work — at all — without our labor.

There’s a second valuable data source in the queries and information we send to the model. Vendors can learn what we want and need, and deep data about our businesses and personal lives, through what we share with AI models. It’s all information that can be used by third parties to sell to us more effectively.

Google’s version of generative AI allows it to answer direct questions from its search engine without pointing you to any external web pages in the process. Whereas we used to permit Google to scrape and index our published work because it would provide us with new audiences, it now continues to scrape our work in order to provide a generated answer to user queries. Websites are still presented underneath, but it’s expected that most users won’t click through. Why would you, when you already have your answer? This is the same dynamic as OpenAI’s ChatGPT: answers are provided without credit or access to the underlying sources.

Some independent publishers are fighting back by de-listing their content from Google entirely. As the blogger and storyteller Tracy Darnell wrote:

I didn’t sign up for Google to own the whole Internet. This isn’t a reasonable thing to put in a privacy policy, nor is it a reasonable thing for a company to do. I am not ok with this.

CodePen co-founder Chris Coyier was blunt:

Google is a portal to the web. Google is an amazing tool for finding relevant websites to go to. That was useful when it was made, and it’s nothing but grown in usefulness. Google should be encouraging and fighting for the open web. But now they’re like, actually we’re just going to suck up your website, put it in a blender with all other websites, and spit out word smoothies for people instead of sending them to your website. Instead.

For small publishers, the model is intolerably extractive. Technical writer Tom Johnson remarked:

With AI, where’s the reward for content creation? What will motivate individual content creators if they no longer are read, but rather feed their content into a massive AI machine?

Larger publishers agree. The New York Times recently banned the use of its content to train AI models. It had previously dropped out of a coalition led by IAC that was trying to jointly negotiate scraping terms with AI vendors, preferring to arrange its own deals on a case-by-case basis. A month earlier, the Associated Press had made its own deal to license its content to OpenAI, giving it a purported first-mover advantage. The terms of the deal are not public.

Questions about copyright — and specifically the unlicensed use of copyrighted material to produce a commercial product — persist. The Authors Guild has written an open letter asking them to license its members’ copyrighted work, which is perhaps a quixotic move: rigid licensing and legal action is likely closer to what’s needed to achieve their hoped-for outcome. Perhaps sensing the business risks inherent in using tools that depend on processing copyrighted work to function, Microsoft has promised to legally defend its customers from copyright claims arising from their use of its AI-powered tools.

Meanwhile, a federal court ruled that AI-generated content cannot, itself, be copyrighted. The US Copyright Office is soliciting comments as it re-evaluates relevant law, presumably encompassing the output of AI models and the processes involved in training them. It remains to be seen whether legislation will change to protect publishers or further enable AI vendors.

So. Who’s making money from AI? It’s mostly the large vendors who have the ability to create giant models and provide API services around them. Those vendors are either backed by venture capital investment firms who hope to see an exponential return on their investment (OpenAI, Midjourney) or publicly-traded multinational tech companies (Google, Microsoft). OpenAI is actually very far from profitability — it lost $540M last year. To break even, the company will need to gain many more customers for its services while spending comparatively little on content to train its models with.

In the face of criticism, some venture capitalists and AI founders have latterly embraced an ideology called effective accelerationism, or e/acc, which advocates for technical and capitalistic progress at all costs, almost on a religious basis:

Technocapital can usher in the next evolution of consciousness, creating unthinkable next-generation lifeforms and silicon-based awareness.

In part, it espouses the idea that we’re on the fringe of building an “artificial general intelligence” that’s as powerful as the human brain — and that we should, because allowing different kinds of consciousness to flourish is a general good. It’s a kooky, extreme idea that serves as marketing for existing AI products. In reality, remember, they are not actually intelligence, and have no ability to reason. But if we’re serving some higher ideal of furthering consciousness on earth and beyond, matters like copyright law and the impact on the environment seem more trivial. It’s a way of re-framing the conversation away from author rights and considering societal impacts on vulnerable communities.

Which brings us to the question of who’s not making money from AI. The answer is people who publish the content and create the information that allow these models to function. Indeed, value is being extracted from these publishers — and the downstream users whose data is being fed into these machines — more than ever before. This, of course, disproportionately affects smaller publishers and underrepresented voices, who need their platforms, audiences, and revenues more than most to survive.

On the internet, the old adage is that if you’re not the customer, you’re the product being sold. When it comes to AI models, we’re all both the customer and the product being sold. We’re providing the raw ingredients and we’re paying for it to be returned to us, laundered for our convenience.

I’m making some experimental updates to my newsletter:

Starting next week, this newsletter will come in several flavors:

Technology, Media, and Society: technology and its impact on the way we live, work, learn, and vote.

Late Stage: personal reflections on living and surviving in the 21st century.

The Outmap: new speculative and contemporary fiction.

Most of Technology, Media, and Society will continue to be posted on this website. I am experimenting with publishing more personal posts and fiction over there.

Prefer to subscribe via RSS? Here’s the feed URL for those posts.

I spent a little time updating my resumé, which is a process that basically sits at the top of all the things I least like to do in the world. This time around I tried to have an eye towards focus: what about the work I do might other organizations find valuable? Or to put it another way: what am I?

I grew up and went to school in the UK. At the time, the A-level system of high school credentials required you to pick a narrow number of subjects to take at 16. In contrast to the US, where university applications are more universal and you don’t pick a degree major until you’ve actually taken courses for a while, British applicants applied for a major at a particular institution. The majors available to you were a function of the A-level subjects you chose to take. In effect, 16 year olds were asked to pick their career track for the rest of their lives.

I now know that I take a kind of liberal arts approach to product and technology leadership. My interests are in how things work, for sure, but more so who they work for. I care about the mechanics of the internet, but I care more about storytelling. I’m at least as interested in how to build an empathetic, inclusive team as I am in any new technology that comes along. The internet, to me, is made of people, and the thing that excites me more than anything else is connecting and empowering them. I’ll do any work necessary to meet their needs - whether it’s programming, storytelling, research, design, team-building, fundraising, or cleaning the kitchen.

Which means, when I picked my A-levels in 1995, and when I applied for universities two years later, that it was hard to put me in a box.

My high school didn’t even offer computing as a subject, so I arranged to take it as an extra subject in my own time. The standardized tests were so archaic that they included tape drives and punchcards. Meanwhile, my interest in storytelling and literature meant that I studied theater alongside more traditional STEM subjects: something that most British universities rejected outright as being too unfocused.

I have an honors degree in computer science but I don’t consider myself to be a computer scientist. I’ve been a senior engineer in multiple companies, but my skillset is more of a technical generalist: technology is one of the things I bring together in service of a human-centered strategy. I like to bring my whole self to work, which also includes a lot of writing, generative brainstorming, and thinking about who we’re helping and how best to go about it.

Even the term human-centered feels opaque. It just means that I describe my goals and the work I do in terms of its impact on people, and like to figure out who those people are. It’s hard to help people if you don’t know who you’re helping. People who say “this is for everyone!” tend to be inventing solutions for problems and people that they only imagine exist. But there’s no cleanly concise way of saying that without using something that sounds like a buzzword.

So when I’m putting together a resumé, I don’t know exactly what to say that ties together who I am and the way I approach my work in a way that someone else can consume. Am I an entrepreneur? I have been, and loved it; I like to bring that energy to organizations I join. A product lead or an engineering manager or a design thinker? Yes, and I’ve done all those jobs. I think those lines are blurry, though, and a really good product lead has a strong insight into both engineering and design. I’ve also worked on digital transformation for media organizations and invested in startups at an accelerator — two of my favorite things I’ve ever done — and where do I put that?

In the end, I wrote:

I’m a technology and product leader with a focus on mission-driven organizations.

I’ve designed and built software that has been used by social movements, non-profits, and Fortune 500 companies. As part of this work, I’ve built strong technology and product team cultures and worked on overall business strategy as a key part of the C-suite. I’ve taught the fundamentals of building a strong organizational culture, design thinking, product design, and strategy to organizations around the world.

I’m excited to work on meaningful projects that make the world better.

I’ve yet to get feedback on this intro — I guess that’s what this post is, in part — but it feels close in a way that isn’t completely obtuse to someone who’s basing their search on a simple job description. It will still turn off a bunch of people who want someone with a more precise career focus than I’ve had, but perhaps those roles are also not a good fit for me.

Perhaps I should be running my own thing again. I promised myself that I would give myself a third run at a startup, and it’s possible that this is the only thing that really fits. At the same time, right now I’m doing contracts, and I love the people and organization I’m working with right now.

If I think of my various hats as an a la carte menu that people can pick from rather than an all-in-one take-it-or-leave-it deal, this kind of work becomes less daunting. Either way, I do think it’s a strength: even if I’m working as one particular facet officially, the others inform the work I’m doing. As I mentioned, I think it’s helpful for an engineering lead to have a product brain, and vice versa. It’s not a bad thing for either to understand design. And every lead needs to understand how to build a strong culture.

But how to wrap all of that neatly up in a bow? I’m still working on it.