Ars on your lunch break: teaching AI to diagnose patient health risks

This week we’re serializing another episode of the After On Podcast here on Ars. Our guest is pediatric oncologist and medical futurist Daniel Kraft. We’ll run the interview in two installments, wrapping it up tomorrow (I mistakenly say it’ll be three installments in my introduction to the audio file—apologies).

Daniel founded and runs the Exponential Medicine Conference, which is one of the largest cross-disciplinary gatherings of life science researchers and innovators.

He also founded and runs the medical faculty at Singularity University—an academic institution so quirky, it could only have sprouted up from Silicon Valley’s soil.

When Daniel does a presentation, he’s the opposite of that speaker we’ve all seen—the one who has to do everything possible to pad their words and slides to fill a time slot. With Daniel, I always sense that there’s an entire presentation lurking behind every slide that he puts on the screen. He just has so much surface area from his two highly complimentary jobs, which connect him to hundreds of startups and researchers every year. Daniel is particularly deep in medical devices—ranging from consumer-grade gear to tools that only turn up in research hospitals. And as an oncologist, he’s of course deeply informed about cancer.

Our conversation is a whirlwind tour of breakthroughs, which may just lurk on life science’s near-ish term horizon. It’s sure to annoy any committed dystopian—so if that’s your perspective, you may want to skip it (but fear not—I’ll present something truly spooky and down-tempo soon enough!).

We begin today discussing Daniel’s background and Singularity University itself. We then delve into the world of advanced quantified-self devices, and how they’re finally starting to link into the caregiving world in ways that could truly save lives.

When I question navigability of the inevitable data glut, Daniel points to the taming potential of AI, by citing some astounding work recently done at Google. By matching 300,000 retinal scans to detailed health histories, their nascent system was able to make remarkably accurate statements about the health and disease risks of scannees—most of which were unrelated to visual wellness. It’s fascinating to imagine what mundane-seeming information—perhaps subtle changes in a person’s gait, or respiratory patterns—might one day become reliable early warnings of disease.

If you enjoy this installment and just can’t wait for part two, you can find the whole shebang in my podcast feed, where it first appeared on May 29th. A full archive of my episodes can be found on my site, or via your favorite podcast app by searching under the words “After On” (the podcast’s title).

If you’re curious about the latest episode in the main After On podcast feed, this week it’s an interview with Yale ornithologist and evolutionary heretic Richard Prum. Rick boldly and brilliant refutes much of the common wisdom about sexual attraction, aesthetics, and more. And the wellspring of his unorthodox ideas is…Charles Darwin himself. If you’re at all interested in the deep roots of human behavior and biology, you’ll find this interview fascinating—even if you’re not at all into birds.

This special edition of the Ars Technicast podcast can be accessed in the following places:

iTunes:
https://itunes.apple.com/us/podcast/the-ars-technicast/id522504024?mt=2 (Might take several hours after publication to appear.)

RSS:
http://arstechnica.libsyn.com/rss

Stitcher
http://www.stitcher.com/podcast/ars-technicast/the-ars-technicast

Libsyn:
http://directory.libsyn.com/shows/view/id/arstechnica

[ufc-fb-comments url="http://www.newyorkmetropolitan.com/tech/ars-on-your-lunch-break-teaching-ai-to-diagnose-patient-health-risks"]

Latest Articles

Related Articles