Product Promotion Network


Alibaba is developing its own driverless cars

The Chinese tech giant comes to the self-driving game later than rivals Baidu and Tencent, but says it will build an entire ecosystem around autonomous cars. The news: Alibaba confirmed today that it’s testing its own self-driving vehicle technology. The effort is led by Gang Wang, a scientist at the company’s AI lab and one of MIT Technology Review‘s 35 Innovators Under 35 in 2017.

Backstory: This isn’t Alibaba’s first dalliance with the auto industry. Early this year, it invested in Xiaopeng Motors, a startup developing electric cars. It also formed a partnership with Chinese carmaker SAIC to build internet-connected vehicles and associated infrastructure.

More than cars: Alibaba says it has bigger ambitions than just robotic taxis. In June 2016, the company launched an AI-powered “city brain” system in Hangzhou, where it’s headquartered, to crunch data from mapping apps and increase traffic efficiency. Simon Hu, the president of Alibaba Cloud, says the firm’s ultimate goal is to produce the kind of autonomous driving that uses data like that so transportation is fully integrated into urban infrastructure.

Why it matters: China is scrambling to compete with America in developing driverless cars as quickly as possible.

This news is another sign that it really means business.

Image credit:

  • Denys Nevozhai | Unsplash

Silicon Valley’s finest all want the Pentagon’s cloud contract, even if it’s not a sensible idea

Some experts worry that awarding the work to a single firm might not get the military what it needs, and it could bring national security issues. The news: The Department of Defense is currently looking to move its digital systems to the cloud–a contract worth as much as £10 billion over the next decade. Defense One writes that, according to one government official, “the race is shaping up as a three-way fight between Amazon, Microsoft, and Google–with Oracle a rather distant fourth.” But: Axios reports that some industry experts are skeptical of the plan, arguing that a single organization might not be able to build exactly what the DoD needs.

Plus, putting all the military’s secretive eggs in one firm’s encrypted basket might be convenient, but could prove hugely problematic if a system was found to be insecure. The Amazon issues: Meanwhile, Oracle’s CEO, Safra Catz, has reportedly told Donald Trump that the process for choosing the cloud provider seems to be rigged in Amazon’s favor. But while Donald Trump apparently told her he wanted the process to be pair, a number of reports suggest that the President plans to “go after” Amazon, which could yet affect the contest.

Now what? It’s still early days in the procurement process, with a final draft of the Pentagon’s wants due to be published May.

But it looks like competition will be fierce and, potentially, uneven.

Image credit:

  • Rudi Riet | Flickr

This neural network examines neurons. Like, the kind in your brain.

As Ernest Rutherford once said, “All science is either physics or stamp collecting.” Well, today’s scientists can feel fortunate that AI is, more and more, being used to keep track of the postage. A new deep-learning system that peers at brain tissue and catalogues individual cells might be the best example yet. The details: Deep-learning algorithms need a lot of data, and in the realm of neuroscience, there’s plenty: cell cultures abound, offering far more brain tissue than the poor interns and low-level researchers who typically do the work could ever hope to sort and label.

Instead, a neuroscientist at UC San Francisco and researchers from Google have teamed up and used the cultures to train a system that could automate some of that tedious work. So far: According to Wired, the algorithm can tell live cells from dead ones and differentiate parts of a cell without the aid of fluorescent labels that human researchers often use (which can damage cells). Why it matters: Beyond simply saving time, automating the process of analyzing samples could speed up drug discovery.

Google also open-sourced the data set and model, which means small labs with fewer resources can put this tech to use as well.

1 2 3 185