How to report better on artificial intelligence - Columbia Journalism Review
- Be skeptical of PR hype
- Question the training data
- Evaluate the model
- Consider downstream harms
- Be skeptical of PR hype
- Question the training data
- Evaluate the model
- Consider downstream harms
Seven principles for journalism in the age of AI
- Be rigorous with your definitions.
- Predict less, explain more.
- Don’t hype things up.
- Focus on the people building AI systems — and the people affected by its release.
- Offer strategic takes on products.
- Emphasize the tradeoffs involved.
- Remember that nothing is inevitable.
When Dan wrote this a week ago, I thought it sounded very far-fetched. Now it sounds almost inevitable.
Well, this is impressive (and brave)—competing a 100 words for 100 days during lockdown …with a baby.
And remember, this isn’t writing and publishing at least 100 words every day; it’s writing and publishing exactly 100 words (that’s the hard part).
I must admit I’ve been wincing a little every time I see a graph with a logarithmic scale in a news article about COVID-19. It takes quite a bit of cognitive work to translate to a linear scale and get the real story.
This might be the most insightful thing that Dan has written since his seminal 2013 Medium article:
The problem with Scrappy Doo, isn’t that he’s annoying, which he is, but that the ghosts suddenly became real, which is an afront to science.
I know this hot-take is about 40 years old, but I’ve been bottling it up.
Digital preservation of dead-tree media:
The Stacks Reader is an online collection of classic journalism and writing about the arts that would otherwise be lost to history. Motivated less by nostalgia than by preservation, The Stacks Reader is a living archive of memorable storytelling—a museum for stories.
A new online publication from Julia Angwin:
Big Tech Is Watching You. We’re Watching Big Tech.
The benchmarks that advertising companies use — intended to measure the number of clicks, sales and downloads that occur after an ad is viewed — are fundamentally misleading. None of these benchmarks distinguish between the selection effect (clicks, purchases and downloads that are happening anyway) and the advertising effect (clicks, purchases and downloads that would not have happened without ads).
It gets worse: the brightest minds of this generation are creating algorithms which only increase the effects of selection.
A terrificly well-written piece on the emperor’s new clothes worn by online advertising. Equal parts economic rigour and Gladwellian anecdata, it’s a joy to read! Kudos to Alana Gillespie for the great translation work (the original article was written in Dutch).
We currently assume that advertising companies always benefit from more data. … But the majority of advertising companies feed their complex algorithms silos full of data even though the practice never delivers the desired result. In the worst case, all that invasion of privacy can even lead to targeting the wrong group of people.
This insight is conspicuously absent from the debate about online privacy. At the moment, we don’t even know whether all this privacy violation works as advertised.
The interaction design of this article is great too—annotations, charts, and more!
This well-researched in-depth piece doesn’t paint a pretty picture for archiving online news:
Of the 21 news organizations in our study, 19 were not taking any protective steps at all to archive their web output. The remaining two lacked formal strategies to ensure that their current practices have the kind of longevity to outlast changes in technology.
A small but perfectly formed progressive web app. It’s a private, offline-first personal journal with no log-in and no server-stored data. You can read about the tech stack behind it:
Your notes are only stored on your device — they’re never sent to a server. You don’t even need to sign-in to use it! It works offline, so you can reflect upon your day on the slow train journey home.
From smart dust and spimes, through to online journaling and social media, to machine learning, big data and digital preservation…
Is the archive where information goes to live forever, or where data goes to die?
Melody Kramer interviews Eric Bailey, formerly of the Boston Globe.
A lovely interactive photo essay charting the results of what happens when evolution produces a life form that allows a planet to take selfies.
Newsvine has closed. Mike reflects on what he built, with a particular eye to the current online news situation.
When we look at how the average person’s news and media diet has changed over the last decade or so, we can trace it directly back to the way these and other modern organizations have begun feeding us our news. Up until 10 or 15 years ago, we essentially drank a protein shake full of news. A good amount of fruits and vegetables, some grains, some dairy, some tofu, and then a little bit of sugar, all blended together. Maybe it wasn’t the tastiest thing in the world but it kept us healthy and reasonably informed. Then, with cable news we created a fruit-only shake for half the population and a vegetable-only shake for the other half. Then with internet news, we deconstructed the shake entirely and let you pick your ingredients, often to your own detriment. And finally, with peer-reinforced, social news networks, we’ve given you the illusion of a balanced diet, but it’s often packed with sugar, carcinogens, and other harmful substances without you ever knowing. And it all tastes great!
There’s also this interesting litmus test for budding entrepreneurs:
We didn’t know for sure if it was going to work, but the day we decided we’d be happy to have tried it even if it failed was the day we ended up quitting our jobs (incidentally, if you are thinking about leaving your job for a new risky thing, this is the acid test I recommend).
The real story in this mess is not the threat that algorithms pose to Amazon shoppers, but the threat that algorithms pose to journalism. By forcing reporters to optimize every story for clicks, not giving them time to check or contextualize their reporting, and requiring them to race to publish follow-on articles on every topic, the clickbait economics of online media encourage carelessness and drama.
A Mozilla-backed project for journalists, publishers, and online communities. The Talk part of it is aiming to fix online comments. The Washington Post is going to try it out.
A wide-ranging post from Andrew on the downsides of Google’s AMP solution.
I don’t agree with all the issues he has with the format itself (in my opinion, the fact that AMP pages can’t have script
elements is a feature, not a bug), but I wholeheartedly concur with his concerns about the AMP cache:
It recklessly devalues the URL
Spot on! And as Andrew points out, in this age of fake news, devaluing the URL is a recipe for disaster.
It’s hard to avoid the idea that the primary objective of AMP is really about hosting publisher content inside the Google ecosystem (as is more obviously the objective of Facebook Instant Articles and Apple News).
A lot has been written about the future of journalism, the importance of businesses like the LA Times being profitable as a way to protect American democracy. I agree with that in theory. But this sort of incompetence and contempt for readers makes me completely uninterested in helping their business.
Like Craig says…
between personal data suction and total disrespect of bandwidth, I'm not sure how you can *not* run ad blockers and browse the web
— A Walkin' Dude (@craigmod) March 26, 2017
It strikes me that Garrett’s site has become a valuable record of the human condition with its mix of two personal stories—one relating to his business and the other relating to his health—both of them communicated clearly through great writing.
Have a read back through the archive and I think you’ll share my admiration.