DataDownload: Why every reporter should be a climate reporter

7 min readJul 24, 2021

DataDownload: Why every reporter should be a climate reporter A weekly summary of all things Media, Data, Emerging Tech View this email in your browser

Some amazing pieces this week.

Kicking off with Buzzfeed climate reporter Zahra Hirji’s exploration of why more journalists aren’t covering climate change. Lack of vetted information at hand, difficulty finding trustworthy sources, intimidating science — plenty of reasons.

Moving on to Melanie Mitchell, who says analogies are essential for AI to map existing knowledge to new situations. Then we dive into Facebook’s never-ending disinformation problems, the inevitable weaponization of app data, and Disney’s growing dominance in the streaming space.

You’ll have a great time watching Neil DeGrasse Tyson discussing the billionaire space race, and listening to a podcast about music’s secret chord. And to wrap up, a deeper look at DeepMind’s decision to open-source hundreds of thousands of protein folding predictions.

Best,
Steve

Steven Rosenbaum
Steve@NYCMedialab.org
Executive Director
The NYC Media Lab Must-Read Congrats, You’re a Climate Reporter Now

In a world battered by Covid-19 and a seemingly endless barrage of brutal wildfires, heatwaves, and floods, has the climate change discussion “gone from being a future thing to a now thing?” According to Buzzfeed climate reporter, Zahra Hirji, the answer is yes. Hirji agrees with Emily Atkin of Heated who recently said on CNN that “every [reporter] should be a climate reporter.”

Journalists on deadline may be reluctant to incorporate the climate change big picture into stories about specific events because they’re intimidated by the science and don’t want to make mistakes. Hirji suggests a possible solution for publishers is to craft “vetted information that [reporters] can slot in” to give readers context and show them that this “isn’t just one random story, it’s part of a trend.”

Buzzfeed recently created constantly updated “context paragraphs” for gun violence stories and is looking at doing the same for climate change. For reporters that want to integrate climate issues into their beat more frequently, “half the job these days is finding who is a trusted source.” Check out the full article for Hirji’s recommended resources.

Galaxy Brain / 18 min read

Melanie Mitchell, author of Artificial Intelligence: A Guide for Thinking Humans, believes that training machines to recognize and process analogies is crucial for unlocking true artificial intelligence. “Today’s state-of-the-art neural networks are very good at certain tasks,” says Mitchell, “but they’re very bad at taking what they’ve learned in one kind of situation and transferring it to another.”

According to Mitchell, this is “something that we humans do all the time without even realizing we’re doing it. We’re swimming in this sea of analogies constantly.” AI models like GPT-3 can “generate language very, very convincingly.” They’re capable of processing vast amounts of information and making data-based predictions without any real “understanding.”

What machines can’t do is “take something [they] already know in some way and map it to something new.” Controversially, Mitchell believes that for AI to become truly intelligent, it will need some kind of body analogous to humans. “My intuition is that we will not be able to get to humanlike analogy [in AI] without some kind of embodiment. Having a body might be essential.”

Quanta Magazine / 10 min read Read More Tech+Media Facebook’s Disinformation Problem Is Harder Than It Looks

President Biden made headlines recently by claiming that Facebook is “killing people” with Covid disinformation on the platform. While Biden walked his comment back somewhat, he pointed out that a recent report identified 12 accounts — dubbed the disinformation dozen — as responsible for the majority of anti-vax propaganda spreading on Facebook.

But that doesn’t mean that “if only Facebook hit its Quit Killing People button, America would be healed again,” said Farhad Manjoo. Cable news outlets like Fox and even elected officials like Marjorie Taylor Greene also play a significant role in spreading anti-vaccine hesitancy and hysteria. What is Facebook supposed to do about that? According to CJR: “Finding the right line between disinformation control, public-health awareness, and outright censorship is not an easy task.”

Columbia Journalism Review / 9 min read Read More The Inevitable Weaponization of App Data Is Here

Giving apps permission to track your location data may make life more convenient — many apps won’t work without it. But that convenience may come at the expense of your privacy. Catholic Substack publication The Pillar recently used location data from Grindr to trace the movements of a high-ranking priest and publicly outed him as gay without his consent, leading to his resignation.

According to Sen. Ron Wyden (D-Or), the dire privacy warnings of experts have come true. “Data brokers and advertising companies have lied to the public, assuring them that the information they collected was anonymous,” Wyden said. “As this awful episode demonstrates, those claims were bogus — individuals can be tracked and identified.”

VICE / 6 min read Read More Disney Is Chipping Away at Netflix’s Dominance.

Netflix’s day in the sun as the undisputed king of streaming video may soon fade to black. The company’s share of worldwide demand interest fell below 50% for the first time in the second quarter of 2021.

According to Parrot Analytics, Netflix’s “lack of new hit original programming and the increased competition from other streamers is going to ultimately have a negative impact on subscriber growth and retention.” Though Netflix founder Reed Hastings claims otherwise, it appears that cheaper services like Disney+ and Apple TV+ are making a dent in the company’s video streaming dominance.

NY Times / 5 min read

“Amazon’s Jeff Bezos is set to be the next billionaire who paid his way into space on his own rocket, drawing intense backlash. Astrophysicist Neil DeGrasse Tyson talks with MSNBC’s Ari Melber about the current billionaire ‘space race,’ and pushes back on critics who argue inequality and other problems on earth should limit private space travel.”

MSNBC (YouTube) / 9 min watch

Watch Now What We’re Listening To Podcast: The Most Beautiful Chord…Ever?

Maybe you’ve heard that there’s a secret chord, but did you know that it’s a borrowed minor four? Hear how artists as diverse as The Beatles, Chopin, Phoebe Bridgers, and Radiohead have used the borrowed minor four chord to unforgettable effect on this illuminating podcast.

Spotify / 36 min listen

Listen Now Virtual Events Free Event: Verge Net Zero
Date: July 27–28
Join more than 5,000 professionals working on sustainability, supply chains, facilities, fleets, manufacturing and other key areas. Register Here.

Free Event: Navigating Startup Law with Becki DeGraw
Date: July 28, 12PM-1PM EDT
Becki DeGraw, partner at Wilson Sonsini, takes a deep dive into practical legal and business issues facing today’s entrepreneurs. Register Here. A Deeper Look DeepMind Creates ‘Transformative’ Map of Human Proteins Drawn by Artificial Intelligence

Last December, DeepMind garnered international attention with its unprecedented performance at the Critical Assessment of protein Structure Prediction (CASP) competition. DeepMind’s protein-folding-prediction model, AlphaFold, was said to have made a leap equivalent to the 2012 ImageNet moment. Andrei Lupas, an evolutionary biologist at the Max Planck Institute for Developmental Biology, said AlphaFold’s results were a game-changer: “This will change medicine. It will change research. It will change bioengineering. It will change everything.”

Now DeepMind is releasing “the most comprehensive map of human proteins” — the AlphaFold predictions of some 350k protein structures. Scientists are likening the potential impact of the data to that of the Human Genome Project: “Most significantly, the release includes predictions for 98 percent of all human proteins, around 20,000 different structures, which are collectively known as the human proteome. It isn’t the first public dataset of human proteins, but it is the most comprehensive and accurate.”

The Verge / 8 min read