Werd I/O

Over on Medium, VP Content Scott Lamb asked:

We’re curious what you think. How do you think Medium should approach AI-generated content? What are good and bad examples of AI content? What are you concerned about? What are you excited about?

Here’s how I replied:

I think my biggest ask is actually on the corpus side of AI writing generation: allow me to prevent my writing from being used as part of an AI system. Companies like OpenAI need to agree to a robots.txt-style system to prevent ingestion that can be broadly used, and then Medium needs to apply it across the board.

Work needs to be done to fingerprint AI writing, but until then, I don't think it can be identified accurately, which means it will always fall through the cracks. Instead, poor quality work - and authors who consistently publish it - should not rise in recommendations.

I wonder if there's a case to be made for creating in-house community-positive AI tools so people aren't using spammy tools from elsewhere? For example, a tool that poses interesting questions and helps an amateur author write more comprehensive original work.

One of the most meaningful pieces of work I’ve been a part of this year was The 19th’s dashboard of what abortion laws look like in every state right now, which has been updated for seven months and counting.

The genesis of the idea came from The 19th’s data visuals reporter Jasmine Mithani, who, with the future of Roe v. Wade in the balance, wanted to provide a go-to way for anyone to see the current state of abortion legislation throughout the US. When Roe was overturned by the Dobbs v. Jackson Supreme Court decision in June, this became vital: sometimes legislation was changing multiple times a day. For people who needed reproductive healthcare or who worked in the space, a resource was badly needed. For citizens and voters in the US, an understanding of how their country was changing off the back of a single court ruling was imperative.

Jasmine built an at-a-glance visualization. The editorial team rallied to continuously-update the page. In product and technology, we sidestepped away from our scheduled roadmap to build tools to more easily update the page, and to support visual elements that didn’t previously exist. We built components that could be re-used later: a toolkit for storytelling nationwide changes like the one we were experiencing.

This kind of work is an example of why I’m proud to work at The 19th. The United States is experiencing a period of unprecedented change, while many of the decisions made here have a profound impact on the rest of the world. Meanwhile, most news is reported by straight, white men, narrowing its lens on a specific demographic. The 19th’s reporters live all over the country and are predominantly women and people of color. (In an organization of over fifty people, I’m one of the only cis white men.) The 19th’s focus on high-quality journalism covering politics and policy through a gender lens has been a largely missing perspective. “You're one of the few publications that reports for me and not just about me,” a reader wrote in recently.

All the reporting at The 19th is made available under a Creative Commons license, and other news outlets are encouraged to republish it for free. That’s why you’ll often see our reporting in places like The Guardian, Teen Vogue, and USA Today. Because The 19th’s lens is unfortunately unique, this republishing policy allows stories that might not be reported elsewhere to find a wider audience. And we’re going to do more: a project I’m working on is to build an open source ecosystem for non-profit software development. Newsrooms do better when they collaborate.

We’re a non-profit startup with a small budget. We don’t have large teams, and nobody is earning VC-funded salaries. Our aim is to make a big impact with a lean operation, and so far it’s been working. We’re also transparent about where our money comes from: there are no anonymous donations. You can read about every single person who has funded us here.

Like other non-profit media, we run seasonal member drives to help expand this group. The ideal is that the majority of our funding should come from small donations from individuals. We’re not there yet - but maybe you can help? Even a recurring donation of $5 makes an enormous difference and helps make news media more diverse. (And, yes, like other non-profit media, if you donate past a certain threshold, you can get some well-designed swag like tote bags.)

Thanks for considering - and for reading. It’s a privilege to work on this problem with this team in the current moment. From the moment it launched, I was glad that The 19th exists - and I’m glad to be on the team.

Lots of people have been worried about deepfakes for a while, but I think the bigger, more pressing concern is detecting AI-generated text.

I’d love to be proven wrong on this hypothesis: the only real market for long-form AI text generation on the web is to generate spam. There are other use cases, for sure, but the people who will be buying and deploying the tech in the short term want to generate huge amounts of content at scale in order to trick people into looking at ads or buying ebooks.

Fingerprinting AI-generated content will allow it to be filtered from search engine results, email inboxes, store listings, and so on. While software providers might not want to remove this content entirely, it seems generally sensible to down-rank it in comparison to human-generated content. Fingerprinting will also be useful in educational settings to prevent AI-generated plagiarism, among other places.

Ironically, the best way to do this might be through AI: what better way to identify neural net output than a neural net itself? While this might lead to false positives, I’m not going to lose a whole lot of sleep about de-ranking content that reads a lot like the output from a software model. The outcome is the same: poor quality, mass produced content is de-emphasized in favor of insightful creativity from real people.

I do think AI has lots of positive uses: for example, I’ve been using DALL-E in my own creative endeavors. It’s a great drafting tool and a way to stimulate ideas. Visual AI tools are avenues for creative expression in their own right. But spam is a problem, and the incentives to create high-volume content for commercial gain are not going away. Previously creating it was human-limited; now it’s CPU-bound. That means any enterprising spammer with a cloud can flood the internet with content as part of an arbitrage scheme. That’s the kind of thing we need to protect ourselves against.

As an employee of a non-profit newsroom, I’m not supposed to do three things: make public partisan statements, donate to political parties or causes, or declare donations. (The latter is why I stopped my long-running Fairness Friday series of posts explaining which social justice cause I’d donated to each week. The donations have continued in private.)

I’m allowed to publicly support movements and advocate for communities, which is why you’ve seen statements from me on trans rights, and you might see me support unions, for example. But most often, I’ll point to links from elsewhere - mostly established news outlets - and simply quote them.

Over time, a picture of my beliefs and ethics certainly emerges. I think even if you’ve only been reading for a week, you’ve probably got a fairly good handle on who I am and what I care about.

Still, I’ve been wondering about listing a set of beliefs, This I Believe-style, specifically to call out my biases and potential blind spots, and also just so you can explicitly know where I’m coming from as a person and filter accordingly. On one hand, it would make it easier for readers to consider anything I write and share objectively, because you’d be more aware of my subjective lens. But on the other, I wonder if that also gives people ammunition to summarily reject an idea that could have merit because they disagree with some other position I hold.

What do you think? Should a blog’s posts stand for themselves, or is it useful to have deeper dives into a person’s belief system?