Mini Briefs

Introducing the First AI Model That Translates 100 Languages Without Relying on English

Facebook AI unveiled a new multilingual machine translation model (MMT) that can translate between pairs of 100 languages without needing intermediary English translations. Prior MMTs often had to translate the source language first into English, and then into the target language. This is because English language data is much more prevalent than data of other languages on the web. To tackle this problem, Facebook developed new data mining techniques that prioritized common language translation pairs, language groups, and “bridge” languages that connect between groups. Together, Facebook was able to collect 7.5 billion sentences for 100 languages, and the trained model can give reasonable results even for language pairs that it was not trained on.

Facial recognition datasets are being widely used despite being taken down due to ethical concerns. Here’s how.

A recent study found that taking down public datasets for training machine learning models due to ethical conerns is not sufficient to curb their uses. For example, the DukeMTMC dataset, which contains recorded faces of people on the Duke University campus, was taken down after a public backslash on how the dataset was popularly used for AI surveillance research. However, more than research 100 papers, many published at prestigious AI conferences, were found to use the dataset after the takedown, and many derivied datasets of DukeMTMC exist and are publicly available.

Together, the availability of the data, and the willingness of researchers and reviewers to allow its use, has made the removal of DukeMTMC only a cosmetic response to ethical concerns.

Podcast

Check out our weekly podcast covering these stories! Website | RSS | iTunes | Spotify | YouTube

News

Advances & Business

• IBM AI model predicts onset of Alzheimer’s disease by analyzing descriptions of a cookie theft - Researchers from IBM and Pfizer trained the model to detect signs of speech impairment in written descriptions of the image, which is frequently used to test for cognitive disorders.

• Microsoft researchers develop assistive eye-tracking AI that works on any device - Gaze tracking has the potential to help people living with motor neuron diseases and disorders extert control over their environment and communicate with others.

• Whisper announces $35M Series B to change hearing aids with AI and subscription model - A few years ago, Whisper president and co-founder Andrew Song was talking to his grandfather about his hearing aids. Even though he spent thousands of dollars on a medical device designed to improve his hearing, and in the process his quality of life, he wasn’t wearing them.

• How Wadhwani AI Uses PyTorch To Empower Cotton Farmers - Cotton is a major fibre crop across the world, cultivated in over 80 countries with nearly 100 million families across the world rely on cotton farming for their livelihood.

• An autonomous construction site is in the works - Retrofitting existing equipment may be a viable path to short-term autonomy on the job site.

• Sam’s Club expands autonomous floor-sweeping robots to all U.S. stores - Sam’s Club will add 372 robot floor scrubbers to the hundreds of units it has already deployed, covering every location in its U.S. chain.

• Applied Intuition Reaches $1.25 Billion Valuation In New Funding For Its Autonomous Vehicle Testing Software - As Tesla, Waymo and others jockey for positioning in the race to make self-driving vehicles mainstream, a lesser-known Bay Area startup has reached “unicorn” status by offering the industry the software needed to run autonomous tests.

• Using AI, Canadian city predicts who might become homeless - TORONTO: As makeshift tent cities spring up across Canada to house rough sleepers who fear using shelters due to Covid-19, one city is leveraging artificial intelligence (AI) to predict which residents risk becoming homeless.

• Facebook uses AI to forecast COVID-19 spread across the US - It could help counties take measures in time to minimize losses.

• Photoshop’s AI neural filters can tweak age and expression with a few clicks - Artificial intelligence is changing the world of image editing and manipulation, and Adobe doesn’t want to be left behind. Today, the company is releasing an update to Photoshop version 22.0 that comes with a host of AI-powered features, some new, some already shared with the public.

• Cyberpunk 2077’s dialogue was lip-synced by AI - Procedural generation provides accurate animations for 10 languages.

• Google announces “hum to search” machine learning music search feature - The company also notes that people need not worry about delivering a perfect pitch to take advantage of this new search capability. On Thursday, Google announced a new “hum to search” feature enabling people to pinpoint songs by simply humming part of a track.

• MIT CSAIL’s AI revives dead languages it hasn’t seen before - Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) claim to have developed a system that can decipher a lost language without needing knowledge of its relation to other languages.

• Kite adds support for 11 new languages to its AI code completion tool - When Kite, the well-funded AI-driven code completion tool, launched in 2019, its technology looked very impressive, but it only supported Python at the time. Earlier this year, it added JavaScript, and today it is launching support for 11 new languages.

• Landing AI launches new visual inspection platform for manufacturers - As companies manufacturer goods, human inspectors review them for defects. Think of a scratch on smartphone glass or a weakness in raw steel that could have an impact downstream when it gets turned into something else.

Concerns & Hype

• Activists Turn Facial Recognition Tools Against the Police - “We’re now approaching the technological threshold where the little guys can do it to the big guys” one researcher said.

• Adobe Unveils Authentication Tool in Battle Against Deepfakes - Adobe Inc. debuted a software tool to help media creators prove their images are real, the latest move by the maker of Photoshop to combat the spread of deepfake technology.

• Automating Image Abuse: deepfake bots on Telegram - Today we go public with the findings of a new Sensity investigation into a newly uncovered deepfake ecosystem on the messaging platform Telegram.

• The true dangers of AI are closer than we think - As long as humans have built machines, we’ve feared the day they could destroy us. Stephen Hawking famously warned that AI could spell an end to civilization. But to many AI researchers, these conversations feel unmoored.

• Tesla is putting “self-driving” in the hands of drivers amid criticism the tech is not ready - Even iPhones have lidar now, but not the cars Tesla hopes to one day make autonomous.

• Researchers find evidence of racial, gender, and socioeconomic bias in chest X-ray classifiers - Research from the University of Toronto, the Vector Institute, and MIT reveals that chest X-ray datasets used to train diagnostic models exhibit imbalance, biasing them against certain gender, socioeconomic, and racial groups.

• Google AI Tech Will Be Used for Virtual Border Wall, CBP Contract Shows - After years of backlash over controversial government work, Google’s technology will be used to aid the Trump administration’s efforts to fortify the U.S.-Mexico border, according to documents related to a federal contract.

• Microsoft and MITRE release framework to help fend off adversarial AI attacks - Microsoft, the nonprofit MITRE Corporation, and 11 organizations including IBM, Nvidia, Airbus, and Bosch today released the Adversarial ML Threat Matrix, an industry-focused open framework designed to help security analysts to detect, respond to, and remediate threats against machine learning systems.

Analysis & Policy

• White House Nears New Rules on Artificial Intelligence - Regulatory guidelines aim to balance fostering growth and controlling industry, administration’s chief technology officer says

• Attention EU regulators: we need more than AI “ethics” to keep us safe - In this post, Access Now and European Digital Rights (EDRi) analyse recent developments in the EU AI debate and explain why we need a bold, bright-line approach that prioritises our fundamental rights.

• It’s time to rethink the legal treatment of robots - In this excerpt from “The Reasonable Robot: Artificial Intelligence and the Law”, Ryan Abbott argues that laws should treat robots more like people. A pandemic is raging with devastating consequences, and long-standing problems with racial bias and political polarization are coming to a head.

• A global collaboration to move artificial intelligence principles to practice - Convened by the MIT Schwarzman College of Computing, the AI Policy Forum will develop frameworks and tools for governments and companies to implement concrete policies.

• Susan Athey: How AI Can Aid in U.S. Economic Recovery - The HAI associate director testified to the House Budget Committee on everything from AI’s falling cost to AI-driven medical applications.’

Explainers

• AI’s next big leap - The unlikely marriage of two major artificial intelligence approaches has given rise to a new hybrid called neurosymbolic AI. It’s taking baby steps toward reasoning like humans and might one day take the wheel in self-driving cars.

That’s all for this week! If you are not subscribed and liked this, feel free to subscribe below!