How Technology Advances the World and Solves Challenges

Emerging Tech Roundup — October 23rd

The Quantious team’s top picks for timely trending news in the tech world.

This week in tech: Instacart and ALDI partner in a pilot grocery delivery program, artificial intelligence makes headway in identifying Alzheimer’s before its onset, Google faces antitrust lawsuit in the US, AR app lets you copy and paste images from reality into designs, Facebook advances hand-tracking through text input capabilities, MIT researchers advance the translation of lost languages.

Instacart and ALDI Partnering to Accept Food Stamps for Grocery Deliveries

(The Verge, October 23)

During the pandemic, grocery and food deliveries services have grown, and there’s been a rise in food assistance in the United States as people work through the challenging economic times. However, many of the available grocery and food delivery platforms don’t accept Supplemental Nutrition Assistance Program (SNAP), otherwise referred to as food stamps. To help with this, grocery delivery platform, Instacart, is launching a pilot program to accept food stamps as payment across Aldi stores in select states. With it, customers can start to buy SNAP-eligible items at Aldi, though delivery fees and tips will not be covered.

The AI That Spots Alzheimer’s From Cookie Drawing

(BBC, October 22)

IBM Research and Pfizer have developed an AI model that can predict whether someone will develop Alzheimer’s disease with up to 70% accuracy, 7 years before onset. Using natural language processing and a “Cookie Theft” cognitive test, the model analyzes a person’s speech as they describe a drawing of children stealing cookies behind their mother’s back in the kitchen. Looking at sentence structure, grammar, and language, the model can help identify cognitive decline. This research is a great step in helping to develop treatments and establish larger, more diverse testing in the future as a way to work toward curing Alzheimer’s.

Google Was Sued by the US Government. Here’s What That Means for You

(CNET, October 22)

Google is being sued by the US Department of Justice in an antitrust case. The Justice Department argues that Google paying billions of dollars a year to be the default option on devices, in addition to “locking in” Google apps across its mobile OS, Android, hurts the market. With this dominance, they claim that Google limits innovation by blocking competition and cornering consumers into using Google’s services. On the other hand, Google argues that its contracts with companies like Apple help to keep prices down for consumers. They claim that Apple would seek to make up revenue elsewhere, like higher-priced iPhones, if they were to no longer receive payments from Google.

Amazing Augmented Reality Copy and Paste App ClipDrop Now Available

(MS Power User, October 22)

A new augmented reality app, ClipDrop, enables users to take objects, people, drawings, and text from reality and drop them into Photoshop or other documents. The app works by using artificial intelligence to detect the object and remove the background, so the resulting image is an isolated object that can be moved over into Photoshop. The app is currently available across Mac, Windows, iOS, and Android for $79.99 a year with unlimited storage.

Facebook Develops Hand Tracking Method to Touch Type in VR Without a Keyboard

(RoadtoVR, October 21)

Facebook’s Reality Labs recently revealed its research into hand tracking, aiming to enable AR/VR users to type via touch — without a physical keyboard. Though basic hand tracking exists today for navigation and gameplay, this next phase of text input required a deep dive into motion research. Since user motion is erratic while typing on a flat surface, researchers needed to isolate both the movement and planned motion of individual fingers to better predict what a user might be intending to type. Although still in the early stages, this is a step forward toward better hand tracking and more advanced AR and VR headsets through a greater understanding of user motion.

Translating Lost Languages Using Machine Learning

(MIT News, October 21)

To decipher lost languages, researchers often turn to machine-translation algorithms and compare to “relative” languages. However, this proves difficult when there’s no dividers, understanding of grammar or vocabulary, and no relative languages to compare to. Although difficult, this work is crucial for understanding the history of the people who used these ancient languages. To help with translations, researchers at MIT’s Computer Science and Artificial Intelligence Laboratory are working on a new system to help decipher lost languages without any knowledge around its relation to other languages. The goal is to be able to decipher these lost languages and connect them to related words in known languages.