During the second day of its annual F8 conference, the head of the social networking company’s mysterious hardware research division talked up the idea of letting people “type” commands or messages by using their minds. Don’t get too excited: it’s not ready to announce a product anytime soon. Meanwhile, it would love for people to try out its new 360-degree digital cameras, built for collecting video that could be used in virtual reality apps

Facebook is working to create a data input method that doesn’t rely on a keyboard, but instead allows the user to type directly from the brain, the company said at its two-day developers conference in San Jose, California.

“In a few years’ time we expect to demonstrate a real-time silent speech system capable of delivering 100 words per minute,” or about five-times faster than a person can type with a smartphone, said vice president of engineering Regina Dugan on the second day of the conference.

Ms. Dugan also heads Facebook’s hardware research unit known as Building eight, which has more than 60 scientists and engineers working on the new keyboardless typing method.

The input method could, for example, allow users to send a text message or email to a friend without taking out a smartphone to type.

Ms. Dugan also tried to calm consumers at the conference, saying the California-based social media giant isn’t interested in detecting a person’s thoughts, but only what the person intends to type.

“We’re not talking about decoding your random thoughts. That might be more than any of us care to know,” she said.

Ms. Dugan referred to research at Stanford University, which has allowed a paralysed woman to type at about eight words per minute directly from her brain.

But the current method requires invasive surgery in which an array of electrodes is implanted to receive data where the brain would normally control the person’s motor functions.

“That simply won’t scale,” Ms. Dugan said, referring the surgery process.

“So we’ll need new non-invasive sensors.”

Facebook is considering wearables such as caps that can read data through the human skull.

While the company may need years to produce a mass-scale device, any advances in the research have potential to be a huge breakthrough in human communications, Ms. Dugan said.

“Even something as simple as a yes-no brain click would fundamentally change our capability.”

Harvard Business School used this piece written by Fasmicro Group Founder for Harvard Business Review  in a recent newsletter. We want you to read again this great piece.

Across Africa, the collapse of the commodity boom has pushed countries and their citizens to invent other ways to survive. The cloud, open source software, cheap computing resources, and ideas from local technology hubs are redesigning the competitive landscape. Could a golden era of entrepreneurship anchored by local innovation be on the horizon? This Harvard Business Review article explores how some African economies are being transformed by the sheer power and optimism of their entrepreneurs. Continue read here.

Amazon released a version of its Alexa speech-recognition service to let developers enable apps for voice. The Lex service comes at a time when chip vendors, including Conexant, NXP and QuickLogic, are hopeful multiple OEMs will start shipping Alexa-like devices this year with their chips.

The news marks the latest foray in a battle of cloud computing giants using machine learning to attract users. It came the same day Facebook was encouraging developers at its annual conference to use new augmented reality capabilities powered by machine learning.

Analysts say Google has an opportunity to marry its superior search service with its Google Assistant, a voice service currently only available on its Pixel smartphones and a new model from LG. Apple is a wild card here with its Siri, an early entrant in the field.

Amazon’s Lex handles speech recognition and natural language processing as a Web service. The company announced support from a handful of early preview users including a bank, an insurance company and a non-profit.

The company launched a reference design in December with software to link to the Alexa service. It already enables Microsoft’s Cortana on some Hewlett-Packard PCs and has had discussions with Google about its Assistant. “I believe in 2018 [voice assistants] will become the largest business for Conexant, said Awsare.

Conexant is working with a handful of OEMs for voice-enabled devices including a digital thermostat maker in Canada. Others want to add voice to set-top boxes, Wi-Fi extenders and even smart light bulbs, he said.

The company already sold nearly 500 of its voice developer kits at $299 each. It expects third-party products at or below the price of an Amazon Dot will hit the market this summer.

Amazon is believed to have sold as many as 7 million Alexa devices to date, mainly using audio chips from Texas Instruments. It claims it will enable 100 million systems for voice by 2020, largely through third-party products