Lightning-Fast AI Chips, Robo-Manta Rays, and the Cost of Cyber Devastation—The Download, April 6, 2017
Three Things You Need to Know Today
Calculating the Cost of Digital Devastation
When a crippling cyber attack hits, who’s going to cover the cost of the damage? Insurers have been known to wildly miscalculate how much cash will be required in the wake of natural disasters, despite the fact that mankind has been living with them for centuries. Cyber attacks, however, are a relatively new threat, but are growing in scale and prevalence—so it's no longer a question of asking if a devastating attack will occur, but rather when. Our own Mike Orcutt investigates how insurers are scrambling to put a price on cyberattack fallout.
Get The Download! Sign up here to have it delivered free to your inbox.
The Economics of Refurbishing Rockets
Recycling spacecraft is savvy—if you can bankroll the R&D. With SpaceX now able to reuse boosters, it’s the hot topic at the 33rd Space Symposium in Colorado Springs. Tory Bruno, CEO of United Launch Alliance—which isn’t reusing rockets—claims the "jury's still out" on the technology, but SpaceX president Gwynne Shotwell says that refurbishment of a booster costs "substantially less than half" as much as using a new one. But the trick requires investment, and Jeff Bezos has revealed how he funds his reusable rocket firm, Blue Origin: by selling $1 billion of Amazon stock every year.
An AI Chip That Stands in for Whole New Data Centers
To handle its burgeoning AI offerings, Google had to design a whole new kind of chip. Now, it’s revealed details of the hardware, called the Tensor Processing Unit, and its performance is impressive. The chip is designed specifically to run pre-trained neural networks, of the kind that translate speech and identify features in images, and they do so 15 to 30 times faster than regular GPUs and CPUs. The effort to build the hardware was clearly worth it: according to Wired, the chip means that Google could avoid building entire new data centers.
Ten Fascinating Things
Depression is a simple-sounding condition with complex origins that aren't fully understood. Here’s how machine learning can help tackle it.
Automation is making manufacturing more efficient, but things can get even better. Big data could save factories more money, energy, and materials.
Turning fossil fuel power plants on and off to cover the intermittency of renewables sounds expensive, but it might not sting as much as we thought.
YouTube TV, the video site’s attempt to offer up traditional viewing to cord cutters, is finally here—but is it worth $35 per month?
Precision medicine is a great idea in the West, but often considered too expensive for the developing world. That may be about to change.
In the sleepy fjord that separates the Norwegian towns of Magerholm and Sykkylven, the peace isn’t even broken by the car ferry—because it’s electric.
IBM has developed a tiny new chip that can help sniff out methane leaks at oil and gas wells around the clock, rather than relying on spot checks.
A new 7-inch robotic manta ray flexes its fins for thrust, swims for 3 hours on a charge, and carries a camera to surveil the sea. Watch it take to the waters.
Here’s a different way to think about liquid-cooling your computer: with jumping droplets of water that dance around on its chips.
There’s a lot that’s good about the World Wide Web. But, Outline asks, why is so much of it hell to use and awful to look at?
Quote of the Day
"When you get a collection of young, nerdy, socially awkward people together and they're emotionally invested into a website you're going to have ... problems.”
— Kevin "Fragmaster" Bowen, once a committed administrator and moderator of the SomethingAwful.com forums, discusses the rise and inevitable fall of the iconic website.
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.