Today, Google Cloud chief Diane Greene announced the company's new push in machine learning and artificial intelligence. There's now a new group in Greene's division that will unify some of the disparate teams that had previously been doing machine learning work across Google's cloud. Two women will take charge of the new team: Fei-Fei Li, who was director of AI at Stanford, and Jia Li, who was previously head of research at Snap, Inc. As Business Insider notes, Jia Li was one of the minds behind the Snapchat feature that lets you attach emoji to real-world objects in your snaps. The news came at the top of a slew of announcements about the product roadmap for Google's cloud services and how they're expanding their use of machine learning, a critical technique for training large-scale AI networks to teach and improve themselves over time. The announcements were all aimed at showing how Google's cloud services include more than just renting time on a server -- that it can provide services to its enterprise customers that are based on its machine learning algorithms.

What happens when you tell two smart computers to talk to each other in secret and task another AI with breaking that conversation? You get one of the coolest experiments in cryptography I've seen in a while. In short, Google Brain researchers have discovered that the AI, when properly tasked, create oddly inhuman cryptographic schemes and that they're better at encrypting than decrypting. The paper, "Learning to protect communications with adversarial neural cryptography," is available here. The rules of the task were simple.

Slack may have unleashed a new demand among the working world to have better tools to chat to each other about their projects and more, but it's also unleashed something else: a torrent of competing products from other tech firms that sell to the enterprise. Today comes the latest in that trend: BroadSoft, a company known for its cloud-based unified communications services, is launching Team-One, a platform for people to chat to each other, bringing in links and data from other projects, and more. Team-One is making its debut today, but it's built on a product that existed before. Earlier this year, BroadSoft acquired a Slack competitor called Intellinote, which it has now integrated with its bigger platform, including its calling and videoconferencing products, adding in new features such as artificial intelligence and bots to help you source data to get your work done. Apart from being a reflection of just how popular collaboration products have become among businesses, the launch of Team-One also another sign of how Slack's early success in this market is getting attacked by competitors from many angles. The startup has seen some of its fast growth slow down, which presents an opportunity for some of those rivals with established customer bases to move in, or for Slack to demonstrate that it definitely is better than the rest.

A month ago, I interview Gary Illyes at Marketing Land and covered this bit at Search Engine Land but it got lost. In short, Gary Illyes from Google said that machine learning and artificial intelligence within the search algorithm can be used to make new ranking signals. He said that Google can use it to say if you combine ranking signal A with ranking signal B, we can make a new ranking signal C that helps improve quality of the search results. Gary Illyes: They are typically used for coming up with new signals and signal aggregations. So basically, let's say that this is a random example and not know if this is real, but let's say that I would want to see if combining PageRank with Panda and whatever else, I don't know, token frequency.

Weather conditions this week will impact the progress of the search for missing Malaysia Airlines Flight MH370, Australian Transport Safety Bureau (ATSB), said Wednesday. The underwater operation to locate the missing plane had already been pushed to end in January 2017 in accordance with the weather forecast. ATSB, which is leading the search for the Boeing 777-200, said in its operational update that weather conditions are unsuitable for both Autonomous Underwater Vehicle (AUV) and Remotely Operated Vehicle (ROV) operations periodically during the week. This week, search vessel Fugro Equator continued underwater search operations in the north of the search area, which is in a remote part of the southern Indian Ocean, using AUV. During the past week, Fugro Equator has undertaken five missions, each one taking an average of 21 hours.

China may be gearing up to pull ahead of the U.S. in the race to overhaul road delivery with fleets of self-driving long-haul trucks. A number of companies are developing automation technologies that promise to lower costs, reduce accidents, and improve overall efficiency for the trucking industry by allowing drivers to make longer trips that include periods of rest. In Europe and the U.S., Volvo, Daimler, Uber, and others are testing trucks capable of driving themselves under expert supervision. But several Chinese-based companies are working on automated trucks, and lenient regulations as well as a desire to overhaul the country's chaotic trucking industry may smooth the way for the technology's introduction. This could provide a handy edge in the race to develop a lucrative new way of hauling goods.

Google's new slew of AI tools includes a kind of game called "Quick, Draw!" that can guess what you're doodling. It's impressive just how good the tool is--though some of its misses would be obvious to the human eye. "This is a game built with machine learning," says a tutorial site. "You draw, and a neural network tries to guess what you're drawing. But the more you play with it, the more it will learn. It's just one example of how you can use machine learning in fun ways."

Facebook's Connectivity Lab today released its high-resolution population maps for Malawi, South Africa, Ghana, Haiti and Sri Lanka, with the promise to make more datasets available over the coming months. The population maps are a joint effort between the Facebook Connectivity Lab, Columbia University and the World Bank, though Facebook is interested in the project as part of its effort to launch wireless communication services in rural regions around the globe. Facebook and friends used software to identify buildings in commercially available satellite images, and then estimated population using census data and a few other surveys and programs. Convolutional neural networks powered a model capable of identifying individual buildings in images from across the world. "There has been a lot of work recently on neural networks that can recognize individual buildings with very high accuracy, but these models are finely tuned on the local characteristics of the region where they are trained," the Connectivity Lab's Tobias Tiecke writes. "We found that these models do not perform well at a global scale with realistic amounts of training data.

What started with Mandarin Chinese is expanding to English; French; German; Japanese; Korean; Portuguese and Turkish, as Google has increased the languages its Neural Machine Translation (NMT) handle. "These represent the native languages of around one-third of the world's population, covering more than 35 percent of all Google Translate queries," according to The Keyword blog. The promise here is that because NMT uses the context of the entire sentence, rather than translating individual words on their own, the results will be more accurate, especially as time goes on, thanks to machine learning. For a comparison of the two methods, check out the GIF embedded below. Google says that the ultimate goal is to have all 103 languages in Translate using machine learning.

A picture may be worth a thousand words, but sound is just as important to how we experience the world as how we see it -- that's why a team at Disney Research is working on a computer vision system that can not only recognize what an image is, but how it sounds, too. In an initial study presented at the European Conference on Computer Vision, the group's system successfully managed to pair appropriate audio with images of doors closing, glasses clinking and vehicles driving down the road. Audio association might be easy for humans, but teaching a computer to do it is actually pretty challenging. Disney researchers trained AI to recognize the sound of images by feeding it a collection of videos demonstrating an object making a specific sound, but background noise, narration or sound made from other objects could easily confuse the system. If the system was fed samples with most of the uncorrelated sounds filtered out, however, it did a pretty good job of suggesting the right sound for each image.