DIY: Tips for Reporting on AI
Pointers from a pro on covering one of the biggest stories in tech
When YR Media launched our Outsmarting AI project, we sat down with journalist extraordinaire Alexis Madrigal, a longtime friend to the organization, to talk about how to develop pitches for stories about Artificial Intelligence. We realized that the gems of insight he dropped back then were too useful to keep for ourselves. So we're sharing them with you! Madrigal offers insight on how to identify potential stories, what questions we need to ask ourselves about our ideas, and how to get the answers we need in the complex, secretive, and sometimes-shady world of AI.
If you’re looking to learn more about AI in general, make your way over to our DIY Outsmarting AI Primer where we define and give examples of different types of AI. You can also hear more from Alexis alongside other prominent figures in AI when you when you check out YR Media's In the Black Mirror: What Artificial Intelligence Means For Race, Art and the Apocalypse.
Coming up with an idea
“I think a lot of it is deciding that what you're noticing in your own Internet world is in fact significant and that you should report on it and that is as important as anything else that other people are reporting on.”
Ask yourself: "What am I seeing or experiencing that I find interesting?" Story ideas about AI are everywhere, including in the apps that you frequently use, your household devices and your social feeds. Start to keep track of these things in a document, note, or notebook.
Turning your idea into a pitch
While the main details about how to create a solid pitch are outlined in DIY: Perfecting the Pitch, Alexis gives us a few helpful questions to refine your AI ideas specifically. In asking yourself these questions, you’re able to surface some of the deeper and sometimes ethical implications of how AI is being used.
Who made it?
How does this work?
What are the unintended consequences of how it works?
Who does it work for?
How else could it work?
Starting your reporting
Once you’ve formulated your story into a pitch, you’ll most likely have to dive into some data and research to get evidence to back up the claims made in your story. This can be especially tricky when looking into stories about AI because technology usually occurs in a mysterious “black box”– where data goes in and some output just ... appears. Your story might involve taking a peek into that black box to see what is going on, and companies and corporations may not be eager to give you that information. Luckily, Alexis walks through how you might go about getting access to what's inside.
TIP 1: "You read a story that's quite like what you're thinking about and you see who they talk to."
Try to locate a story or even a few stories that are similar to what you want to report on. Take a look at who they talk to and the sources they use. See if you can get some insight into who can (and is willing to) give you some intel. Oftentimes, people at companies cannot and will not disclose sensitive information; however, you might be able to track down a prior employee or researcher who is willing to give you the answers you need.
TIP 2: "Hey, let's just get to the facts here."
You may have very good reasons to believe that a company’s AI use is corrupt or unjust; however, it’s a good idea to approach your contact as if you haven’t already made up your mind about what is going on, so that they are not on the defensive — and so you stay open to what you learn. A good way to ask for information “I’ve noticed …” and “Can you walk me through…”
TIP 3: "Get your own set of data and go to them and you say, 'Hey listen, we have this thing here. We have something.'"
When you’re pursuing a story in AI, an especially strong option is to develop a set of data based on your personal research. This is a good move for a few reasons: Assembling your own data enables you to be selective and precise about exactly what information you’re gathering for your story, and you’ll be able to take this dataset to your contact or interviewee and use it to ask specific questions based off of indisputable observations.
If you’re curious about how to gather data and don’t know where to start you can check out DIY: Telling Stories with Data.
The implications that AI and technology have for our society are huge. You don’t have to look far when you’re figuring out which stories to tell. Alexis left us with a few inspiring parting words for young reporters that should give you all the confidence you need to pitch your AI story.