Three takes on a complicated question

In a recent episode of the TDS Podcast, guest Rosie Campbell, Head of Safety Critical AI at Partnership on AI (PAI), chatted with Jeremie Harris about the potential risks of publishing cutting-edge AI research. I admit that It’s a problem I never thought about before listening to their conversation; for one, I’m most definitely not an AI researcher, so I never had to.

A deeper reason, however, is the mental image I’ve constructed over the years of what dangerous AI might look like. Popular culture has conditioned me (and, I suspect, many of you, too) to imagine either a sinister…


Taking a step back from the hum of day-to-day model-training and number-crunching, Phoebe Wong asks some very big questions in her 2019 post about the types of knowledge data scientists explore — and generate. Starting off with the long-simmering tension between models that predict real-world outcomes and those that seek to explain the causal forces at play, Phoebe suggests that we might not be in either/or territory after all: “predictive and causal models serve very different purposes and require very different data and statistical modeling processes, and often we need to do both.”

The debate around advanced degrees is as lively as ever

Photo by Haneen Krimly on Unsplash

Common perceptions of data science (and of tech more broadly) suggest it is an egalitarian, merit-based field—one where it doesn’t matter whether you graduated from Stanford or aced a few online courses. Show employers you know your stuff and can help them grow, and an intellectually rewarding, well-paying job is within reach.

The reality is often more complicated, of course. As thousands of antsy blog posts and Reddit threads show, the job market in data science-related fields is extremely competitive, and the required-skills list for some roles can seem daunting even for seasoned pros. Throw in a dose of pandemic-fuelled…

An accessible, illustrated introduction to a complex topic

In recent years, GANs (generative adversarial networks) have been all the rage in the field of deep-learning generative models, leaving VAEs in relative obscurity. But there’s much to gain from a solid footing in variational autoencoders, which tackle similar challenges but use a different architectural foundation. If you were looking for an engaging, accessible way to learn more about VAEs, Joseph and Baptiste Rocca’s introduction hits the spot. They define terms, walk us through the various elements that make up VAEs and how they relate to each other, and add beautiful illustrations for all the visual learners out there.

Reading List

Discover some of the best Twitter data analyses from the TDS archive.

Image by author, who was momentarily not on Twitter at the time it was taken.

Anecdotally, it would seem that you can be on Twitter, or you can attempt to find joy, health, and balance in your life, but not both. The writers, journalists, and niche food-opinion-havers in my timeline form a fairly diverse lineup. What unites us is a strong ambivalence—yes, I’m mincing words here—about the space itself: the “hellsite” we can’t leave behind.

Digging into the TDS archive these past few weeks has had the unexpected effect of suggesting a different possibility—an alternate reality, even. Here were dozens of data scientists and AI experts spending massive amounts of time on Twitter and… doing…

Or is it “chefly” or “cheffily” or “chuffingly”

A cityscape at night, showing a lit stadium, hi-rises, and a body of water
A cityscape at night, showing a lit stadium, hi-rises, and a body of water
Image by author, who enjoys walking at night

According to some North American etiquette authorities, it’s acceptable to send newlyweds a gift up to an entire year after the wedding. I remember being floored by this tidbit when I first learned about it many, many years ago, as I grew up in a culture where it would be more or less unthinkable for an adult to show up empty-handed at any social gathering, let alone a wedding. A year! What if… they move to an off-the-grid cabin where your fancy cocktail glasses might never reach them? What if a global pandemic starts just as you bestow the couple…

Ben Huberman

Editor in Chief, Towards Data Science. Previously: Editorial lead, Automattic & Senior Editor, Longreads.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store