Social Media Gains Ground as Source for News
According to a new report published by the Pew Research Center, nearly 20% of all U.S. adults get their news from social media. That’s just 1% higher than those who rely on print sources, an embattled medium that seems to have more lives than the average housecat.
Still the go-to, television continues to attract the biggest audience (49%) despite steady declines in viewership. Among the three primary types of news programming, local remains the most popular (37%), followed by cable (30%) and national (25%).
When asked whether they got their news fix from a streaming device, 9% of Pew’s respondents said they did so regularly. Not surprisingly, there were variances between the different age groups: those 65+ are 5 times more likely to get their news from the ol’ tube versus those than those aged 18 to 29.
Alexa Gets AI Upgrade
With the help of some rather sophisticated AI, Amazon’s Alexa will soon be able to tell whether a query for “Mary Poppins” refers to the novel, the soundtrack, or the film.
In his recent blog post, Chengwei Su, a senior applied scientist in Alexa’s AI Natural Understanding group, detailed a system that allows the AI models responsible for Alexa’s core domains (e.g., books, music, movies, etc.) to improve in accuracy independent of one another.
Su, who plans to present his research at the IEEE Spoken Language Technologies Conference in Greece next week, says the work (which should excite voice search experts, enthusiasts and brands whose products exist in a variety of formats) is already in production.
In related news, Amazon debuted Alexa Answers last week, a feature that lets customers tackle uncommon questions by submitting answers that may be distributed to millions of Alexa users around the world.
Bose Enters Wearables Market
Bose, the company best known for its high-end audio products, has begun taking preorders for Frames, a new wearable that’s being positioned as “audio sunglasses.”
Frames will “use integrated speakers to give users location-specific sonic cues as they navigate environments,” writes VentureBeat, and unlike Google’s failed Glass, will look more like a pair of “typical sunglasses” than some Star Trek prop.
While not as sophisticated as a visual AR headset, Frames will feature Bluetooth connectivity for 3.5 hours of battery-powered audio playback when paired with one’s phone. They’ll also include a nine-axis head motion sensor to determine the direction a user is facing which, when paired with a phone’s GPS, will provide location services and enable “audio AR” apps that will know, for example, that you’re in a museum and looking at a specific work of art before delivering information about what you’re seeing.
Bose will launch Frames in January, but without AR support. They expect to provide an update on their AR progress in March when they hit the SXSW circuit.
Instagram Unveils New Voice Messaging Features
Following months of not-so-quiet testing, Instagram has unveiled its new voice messaging service.
Part of Instagram Direct, their “walkie-talkie feature” is designed to give users a(nother) way to easily record, send and receive minute-long voice messages, one-on-one or via group chats.
Akin to products already available on platforms like Viber, Zello, Telegram, Facebook, WhatsApp and LinkedIn, Instagram’s offering does include one interesting nuance: hands-free direct messaging, which could make it – as TechCrunch writes – “a more appealing chat app for drivers, people on the move with their hands full or users in the developing world who want a more intimate connection without having to pay for the data for long audio or video calls.”
Google Lens Functionality Added to Google’s iOS App
Mark our words: the addition of Google Lens functionality to Google’s iOS app will expedite the rise of visual search and 2019 will be year it all happens.
Released this week, the feature enables the visual search of products, buildings, plants, animals, QR and barcodes, business cards and virtually anything that features text in English, French, German, Italian, Korean, Portuguese and Spanish.
Eventually, this technology could be used to identify interest in a service, product, or place (like the Empire State Building, for example) before serving an ad.
It could even make product images in ads shoppable, all without needing a physical ad unit.