MIT Technology Review

The discovery may explain how these buildings have survived for so long in earthquake zones.

Taking down the ‘drunk ’ Pelosi video could set a precedent for censoring political satire or dissent.

A look at historical case studies shows us how we handle the liability of automated systems.

Can the world’s great powers agree on rules of the road for artificial intelligence?

And how it might even be a good thing.

This week’s most thought-provoking papers from the Physics arXiv.

Trade barriers and immigration controls might lead different countries to adopt incompatible products, impeding global innovation.

We are getting a clearer picture of what the social network is working on behind the scenes: Globalcoin

The idea of driving on an inverted track has attracted lots of analysis, but a previously overlooked approach could let an upside-down driver keep going safely and indefinitely.

Here’s a rundown of all the major news in the past week

The long-promised overhaul that will unlock what creator Vitalik Buterin calls the “world computer” needs a cash infusion.

Novartis will sell the world’s most expensive drug, a treatment called Zolgensma to treat spinal muscular atrophy.

The apparent spike in CFC-11 production in China underscores the importance of ongoing emissions monitoring—and the limits of our current system.

An MIT robot collaborates with a person by tracking his or her muscles.

To get back from the lunar surface, the astronauts had to rendezvous in lunar orbit. Sheila Thibeault helped make that possible.

By holding their money in “stable” cryptocurrency, Argentines might avoid the notorious volatility of their peso. First, though, they have to be persuaded.

Researchers trained a neural network to map audio “voiceprints” from one language to another.

The Air Force Artificial Intelligence Incubator aims to develop technologies that serve the “public good,” not weapons development.

Sucking methane from the air might deliver a bigger bang for the buck than just removing carbon dioxide.

New research could make deep-learning models much harder to manipulate in harmful ways.

This week’s most thought-provoking papers from the Physics arXiv.

It might have been less exuberant than last year, but crypto hype isn’t going away anytime soon—and there’s still clearly big money to be made

Dietary habits are notoriously difficult to monitor. Now data scientists have analyzed sales figures from London’s biggest grocer to link eating patterns with local rates of high blood pressure, high cholesterol, and high blood sugar.

Facebook, Palantir, Microsoft, and more see big money in migration caused by global warming—and they’re betting on everything from biometrics to blockchain.

The city government can’t use the technology, but private companies still can, and regulating those uses is a thornier problem.

No solar material has managed to supplant silicon. Perovskites, which are far cheaper and can be made into flexible modules, could change that.

He taunted the health authorities. Now he stands accused of pretending to be one.

If you have difficulty hearing, it can be tricky to make out a single speaker in a noisy room. A system that amplifies the voice you want to listen to could help.

By creating a new genome, scientists could create organisms tailored to produce desirable compounds.

Here’s a checklist for assessing the quality and validity of a company’s machine-learning product.