Discover more from Last Week in AI

Weekly text and audio summaries of the most interesting AI news, as well as editorials commenting on recent events

Over 23,000 subscribers

Last Week in AI #96

Invisible workers of AI, how AI can help with dyslexia, and more!

Last Week in AI

Dec 21, 2020

Tabitha Goldstaub with a laughing smile sitting in a comfy chair with open patio doors to a garden behind her — Image Credits: Mollie Rose/The Observer via The Gurdian

Happy Holidays! Last Week in AI will take a break over the next two weeks and resume in January. We wish everyone a fantastic holiday season and will see you in the new year!

Mini Briefs

AI needs to face up to its invisible-worker problem

In an interview with the MIT Tech Review, Saiph Savage, director of the Human-Computer Interaction Lab at West Virginia University, reviews her work on addressing the "invisible-worker" problem of AI. In the U.S., more than a million people work on platforms like Amazon Mechanical Turk, which tasks human workers to label data used later to train machine learning models. Other tasks, like transcribing audio for voice assistants and flagging inappropriate content, are also common.

Savage points out the main problems with these platforms are the low wages and lack of career development opportunities. While more than 250k workers in the U.S. earn at least three-quarters of their income through such work, they're only paid on average $2 per hour, far below minimum wage. Programs that help workers optimize for task selection, share advice, or allow workers to run their own data-labeling platforms may help alleviate these issues.

How artificial intelligence helped me overcome my dyslexia

Tabitha Goldstaub, co-founder of a "market intelligence platform for AI," shares her experience with how simple AI-powered tools help her tackle challenges caused by dyslexia. Improved spelling and grammar checking software, like SwiftKey and Grammerly, have improved her ability to quickly communicate, and so do recent advances in voice-to-text. An important aspect of AI-based tools is their ability to adapt to user data over time, improving their effectiveness the more they're used. Goldstaub emphasizes that the usefulness of AI-powered automation technology means that they will only grow more popular and powerful in the future, and it is in the interest of both individuals and society as a whole to adapt accordingly to mitigate the disruption AI will have on jobs and its risks in biased decision making.

Podcast

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

News

Advances & Business

Has AI adoption plateaued, or is it just catching its breath? - "It has been a wild year in every quarter, and AI development is no exception. On the whole, the year has been mixed for AI, as there have been both notable advances and new revelations about abusive applications of the technology."
AI test rules out a COVID-19 diagnosis within one hour in emergency departments - "The rapid screening test has been developed by researchers at the University of Oxford"
Best week ever for robotaxis? - "This has arguably been the best week ever for the development of autonomous vehicles. Robotaxis, in particular, marked new milestones."
In a first, Air Force uses artificial intelligence aboard military jet - "The Air Force allowed an artificial-intelligence algorithm to control sensor and navigation systems on a U2 Dragon Lady spy plane in a training flight Tuesday, officials said, marking what is believed to be the first known use of AI onboard a U.S. military aircraft."
Robot dogs could someday explore deep caves on Mars - "Mars exploration is going to the dogs. The robot dogs, that is. In a presentation on Dec."
Artificial Intelligence Discovers Surprising Patterns in Earth's Biological Mass Extinctions - "The idea that mass extinctions allow many new types of species to evolve is a central concept in evolution, but a new study using artificial intelligence to examine the fossil record finds this is rarely true, and there must be another explanation."
AI-Enabled Cheetos Offer Promise of the Perfect Puff - "PepsiCo has developed an AI system to ensure each Cheetos meets its standards on shape, density and crunchiness"
Deepdub uses AI to localize voice dubbing for foreign language films - "Deepdub is launching its platform that uses AI to dub voices into videos so entertainment companies can localize content in multiple languages at the click of a button."
Modulate's ToxMod uses AI to scan game voice chat for toxic speech - "Today, Modulate has launched ToxMod, a new service that uses AI to scan voice chat in video games for toxic speech or other bad behavior."

Concerns & Hype

EU rights watchdog warns of pitfalls in use of AI - "The European Union's rights watchdog has warned of the risks of using artificial intelligence in predictive policing, medical diagnoses and targeted advertising as the bloc mulls rules next year to address the challenges posed by the technology."
"It's really on them to learn": How the rapid rollout of AI tools has fueled frustration among clinicians - "Health workers are increasingly being asked to incorporate AI into their workflow, but say they lack the data needed to trust the tools."
DeepMind A.I. unit lost $649 million last year and had a $1.5 billion debt waived by Alphabet - "Losses at DeepMind, the artificial intelligence firm owned by Google parent Alphabet, are continuing to grow."
Climate change falls down the agenda at A.I. lab DeepMind after energy heads leave - "The DeepMind Energy unit, which has attracted its fair share of attention over the years, has vanished and none of the company's staff mention it on their LinkedIn profiles."
Tensions in Google's ethical AI group increase as it sends demands to CEO - "Tension between Google's Ethical AI group and executives at the company rose on Wednesday, as employees sent a list of demands regarding the recent departure of one of the team's leaders - a prominent Black woman in a field that's largely White and male."
"I started crying": Inside Timnit Gebru's last days at Google - and what happens next - "By now, we've all heard some version of the story. On December 2, after a protracted disagreement over the release of a research paper, Google forced out its ethical AI co-lead, Timnit Gebru."

Analysis & Policy

8 Leading Women In The Field Of AI - "Yet there are many brilliant women at the forefront of AI today. As entrepreneurs, academic researchers, industry executives, venture capitalists and more, these women are shaping the future of artificial intelligence."
Shaping the Development of Trustworthy AI - "As various artificial intelligence systems developed so far carry significant risks, the EU has emerged as a key player to define and shape trustworthy AI."
Network effect: visualizing AI connections in the natural sciences - "Collaborations on AI-related papers in journals tracked by the Nature Index reveal country strengths."

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