Inside AI - December 12th, 2019

Inside AI (Dec 12th, 2019)

Subscribe | View in browser

1. At today's NeurIPS conference, Facebook launched its deepfake detection challenge, which asks researchers to build open-source detection tools that spot deepfake videos. The challenge includes a data set, which contains 100,000-plus videos that were specially created from scratch to help with deepfake research. Participants can use the data set to create new models and will receive scores, according to a Facebook AI blog post. Facebook says it's earmarked more than $10 million in awards and grants for entrants, who have until March to submit. The challenge is being done in conjunction with Amazon Web Services (AWS), Microsoft, the Partnership on AI, as well as academics from eight universities: Cornell Tech, MIT, University of Oxford, UC Berkeley, University of Maryland, College Park, and State University of New York at Albany. For its part, AWS says it will contribute $1 million in service credits and host entrants’ models. - IEEE SPECTRUM

2. AI expert and professor Noel Sharkey is asking the U.K. to issue a moratorium on all decision algorithms, which he said are currently "infected with bias." The professor of AI and robotics at the University of Sheffield told The Guardian that the country should halt all algorithms that have an "impact on people’s lives" because "they are not working and have been shown to be biased across the board.” While companies like Google and Microsoft have been working on a solution, so far none has been found, he noted. Until that happens, Sharkey argued that researchers should test the AI systems on hundreds of thousands or (ideally) millions of people - what he described as "large-scale pharmaceutical-style testing" - to eliminate these in-built biases. - THE GUARDIAN

3. Researchers at Kneron tricked facial-recognition software at an airport and other locations, according to the San Diego-based AI firm. Kneron says it used a printed mask of another person's face to fool the payment tablets, which were run by the Chinese companies Alipay and WeChat, as well as facial recognition software at the Chinese border checkpoint and a passport-control gate at Amsterdam's Schiphol Airport. However, they were unable to fool Apple's Face ID or Huawei's system, which both use structured light imaging, they said. Kneron CEO Albert Liu said companies "should be held accountable if they do not safeguard users to the highest standards." - BUSINESS INSIDER

4. Publications of AI-related papers rose by 300 percent between 1998 and 2018, according to the latest AI Index report. The report's creators released a tool to search through the AI research papers and another to help readers sift through country-level data on AI research and investment. The survey, which was created by researchers at Harvard, Stanford, OpenAI, and other organizations, had generally positive things to say about various and current AI trends, which we've highlighted below:

  • The U.S. is the global leader in AI by most metrics, with slightly below $12 billion in private AI investment compared to China's $6.8 billion.
  • The time required to train a machine vision algorithm on ImageNet fell from about three hours in October 2017 to 88 seconds in July 2019.
  • NeurIPS anticipates hosting 13,500 attendees this year, which is up 800 percent from 2012.
  • More than 21 percent of computer science PhDs choose to specialize in AI.
  • Nearly 10 percent of global private investment went into autonomous vehicles, amounting to $7.7 billion. - THE VERGE

5. There were no winners in Microsoft's MineRL challenge, which asked coders to create AI agents that could locate gems in Minecraft. Of the 660 entries submitted, no one fully solved the task, according to Katja Hofmann, a principal researcher at Microsoft Research. Entrants were allowed to use only a single GPU and four days of training time for their AIs. Hofmann said the task posed was "very hard" and entrants "have made a lot of progress and learned to make many of the tools needed along the way." The results are formally scheduled to be announced on Saturday at the NeurIPS AI conference in Vancouver. - BBC

6. NASA officials said they plan to use machine learning to analyze data collected by future telescopes, which could help in the search for extraplanetary life and asteroids. AI tech could help analyze large data sets collected from tools like the James Webb Space Telescope or the Transiting Exoplanet Survey Satellite (TESS) mission, according to the agency. Giada Arney, an astrobiologist at NASA's Goddard Space Flight Center, said the data from these observations will be "sparse and noisy," making it "really hard to understand." In this regard, AI "has so much potential to help us," she said. - SPACE.COM

7. PrecisionHawk, an AI-based platform to analyze drone-sourced data, landed $32 million from investors including Millennium Technology Value Partners, Eastward Capital Partners, and Third Point Ventures. The Raleigh-based company will use the fresh capital to grow its software products, expand marketing and sales, and hire more software engineers, CEO Michael Chasen told The News & Observer. “We have seen 100 percent growth for the last two years,” said Chase. PrecisionHawk has a fleet of drones and sensor hardware, flight and analytics software, and other drone-related services for businesses, which it uses to help clients choose the technology and develop with flight plans. Its jobs board matches drone pilots with clients, who have access to a platform that uses machine vision to create maps, models, and measurements based on data. - VENTURE BEAT

8. Half of companies spend between eight and 90 days deploying a single machine learning model, according to a new report from Algorithmia. The company released its 2020 State of Enterprise Machine Learning study, which surveyed 750 business decision makers at various organizations. According to Tech Republic, there were seven key findings, which include a rise in data science for machine learning - which will increase the demand for data scientists - and a long road to deployment, meaning machine learning projects are so new that data scientists aren't as familiar as they could be with the process right now. The report found that nearly 20 percent of companies polled take longer than 90 days to deploy a single model. - VENTURE BEAT

9. AX Semantics - which developed AI software that creates human-like written content - launched in the U.S. today. The Stuttgart, Germany-based company opened an office in Sunnyvale, California, which is in Silicon Valley. The company told Forbes that its natural language generation platform can produce written content in more than 110 languages, filling up entire websites with things like earnings and weather reports and sports scores. The company counts Deloitte, Porsche, and Nestlé among its customers. - FORBES

10. Mashable is selling an AI-powered electronic chessboard at 27 percent off right now. The Square Off board uses automated technology to make its own moves, competing against human players at 20 difficulty levels. According to Mashable, it typically retails for $389 but is on sale for $329.99 and buyers can save an extra 15 percent with the code MERRYSAVE15, lowering the price to $280.50. - MASHABLE

Written and curated by Beth Duckett, a former reporter for The Arizona Republic who wrote a book about the solar industry and frequently covers hobby and commercial drones. You can follow her tweets about artificial intelligence news here.

Edited by Sheena Vasani, Inside Dev editor.

Copyright © 2020, All rights reserved.

Our mailing address is:
767 Bryant St. #203
San Francisco, CA 94107

Did someone forward this email to you? Head over to to get your very own free subscription!

You received this email because you subscribed to Inside AI. Click here to unsubscribe from Inside AI list or manage your subscriptions.

Subscribe to Inside AI