Inside AI - December 23rd, 2019 |

Inside AI (Dec 23rd, 2019)

SF amends facial recognition ban / AI-powered shopping carts / The Guardian's 'Zuckerbot'

Subscribe to Inside AI

Inside AI. ###PIXEL:IMG###
Subscribe | View in browser

1. San Francisco has amended its facial recognition ban in local government to allow iPhones with Face ID, Wired reports. As part of the new amendments, SF city agencies can purchase iPhones and other products with facial recognition, as long as other features in the products are considered "critically necessary" and no alternatives exist. City workers are still barred from actually using Face ID and have to type in passcodes instead. After passing the facial recognition ban in May, San Francisco learned quickly that the city already owned "a lot of facial recognition technology," including the iPhones for city employees, which forced it to amend its law, according to Wired. As other cities follow suit with similar bans, many are including exemptions for personal devices with built-in facial recognition and tagging features. - ENGADGET

2. AI-powered shopping carts could compete with store-placed cameras in the race for automated checkout services, CNN reports. Caper and Veeve, which are two startups developing the technology, claim that it's easier to add cameras and sensors to shopping carts rather than placing them around an entire store, as is the case with Amazon Go stores. The technology to run a Go store – which currently spans around 2,000 square feet or less – is too costly to use in a large grocery store, argues Ahmed Beshry, co-founder of Caper. The company's technology used AI to discern what items a customer places in a car, while a built-in scale weighs the products like fruits and vegetables. Once a customer pays, lights on the shopping cart which turn green or red indicate whether the order is complete. - CNN

3. The Guardian worked with Botnick Studios to create an AI-powered bot in the likeness of Facebook's Mark Zuckerberg. The "Zuckerbot" is a neural network trained on more than 200,000 words from the CEO's speeches, interviews, and blogposts, according to the Guardian. Zuckerberg has refused to grant interviews to the publication, which broke the initial story about the social media giant's Cambridge Analytica scandal along with The Observer. When asked about the purpose of Facebook, here's what Zuckerbot had to say: "First, I want to thank you all for being a part of my mission. The purpose of Facebook is to bring about internet for people in the digital room. It is about advertising dollars to people who pay to be able to pay. We believe in building $8bn of voice on the platform for dads who are hungry for coffee." - BOING BOING

4. Cornell University researchers used AI to analyze images of food and drink products purchased by Oktoberfest-goers in Germany. The findings show that machine learning for detecting objects can be used in automated checkout services, but the detection results must be stable. As part of the research, the researchers suggested a system they called "pseudo labeling" - or using a model trained on already-labeled data to annotate unlabeled data - to work around the labor-intensive process of analyzing massive amounts of data. The published datasets are available online. - ZDNET

5. More than 70 percent of people polled in a recent survey said they are open to restaurants incorporating AI into their services. The poll of more than 2,000 U.S. adults, done by The Harris Poll, focused on customer sentiment in quick-service restaurants and fast-casual restaurants. In the AI category, 43 percent of respondents said they are most interested in AI driving down the cost of menu items, while 42 percent said they would like AI to speed up ordering. Sixty-six percent said they would be open to using a touchscreen to order, higher than the 42 percent who preferred voice ordering. - THE SPOON

6. Eddy Travels, which developed an AI bot that can suggest flight deals, says it plans to expand into car rentals, hotels, and other travel services. The year-old company recently closed a $500,000 pre-seed round led by Techstars Toronto, Practica Capital, and Open Circle Capital VC funds, as well as angel investors from the U.S., the U.K., and Canada. The company's travel assistant bot reads people's text and voice messages and translates that into personalized suggestions for flights. Eddy processes about 40,000 flight searches a month. - TECHCRUNCH

7. GM is in discussions with the National Highway Traffic Safety Administration (NHTSA) to possibly allow the operation of autonomous vehicles without steering wheels and other controls. GM has been petitioning the agency to allow it to operate vehicles without these controls since January 2018, which was later followed by a petition from delivery startup Nuro to allow driverless delivery vehicles without windshields. Acting NHTSA Administrator James Owens said his agency will make a decision soon on the issue. - REUTERS

A version of this story first appeared in Inside Auto.

8. Tencent has released a paper detailing how its AI defeated over 99 percent of human opponents in the multiplayer online battle game known as Honor of Kings. The paper's co-authors describe how Tencent’s technique is highly scalable and could help developers create other similar systems for problem solving. Tencent’s architecture uses reinforcement learning learner, AI server, dispatch module, and memory pool. For each day of training, the system "accumulated the equivalent of 500 years of human experience." - VENTURE BEAT

9. Perfect Corp., which created the YouCam Makeup virtual reality app, is using its AI technology to power a hair color app for Sally Beauty. The ColorView virtual hair color experience is now on the Sally Beauty app and in 5000 stores, according to the company. The AI-based YouCam technology aggregates hair color and makeup recommendations in Sally products and allows users to try out different shades. - RETAIL CUSTOMER EXPERIENCE

10. Fast Company interviewed venture capitalists, tech execs, and analysts about their thoughts on technology trends for 2020, including what's in store for AI. The publication predicts that AI's role in business will grow and people will continue to become more aware of AI, including the problems posed by deepfakes and privacy issues. We highlighted some of the responses below:

"AI has the potential to democratize healthcare. With a natural place in virtually all areas of care, from prevention to diagnosis to treatment, it can lower costs, provide greater access, and give everyone the very best doctor, leveling the playing field on a global scale." - Vijay Pande, general partner, Andreessen Horowitz

"As many have feared, in 2020 we’ll see the first malicious use of deepfakes and other forms of synthetic media with the aim of influencing the presidential election. Though there will be at least one attempt that does initially cause a good deal of outrage, these efforts will largely fall flat." - Peter Rojas, partner, Betaworks Ventures

"With the advancement in voice technologies and with the maturity of chatbots and custom digital assistants coming into the market, voice will bring a two-way conversation in 2020." - Kuldip Pabla, senior VP of engineering, K4Connect

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 the latest news in artificial intelligence here.

Copyright © 2021, 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