New robot from Boston Dynamics, Deepfake Amazon workers, and more!

Last Week in AI #110

Apr 5, 2021

Image Credits: Boston Dynamics via The Verge

Mini Briefs

Boston Dynamics unveils Stretch: a new robot designed to move boxes in warehouses

Boston Dynamics recently released a demo for its newest robot Stretch, a mobile robot with a large arm that can move boxes around in a warehouse. The company reports that Stretch's cleverly designed counterweight system hidden in its base allows it to quickly swing its arms around, moving up to 800 boxes per hour. The mobile base also allows users to deploy the robot without installing permenant robot mounts or additional hardware. Stretch has a battery life of up to 8 hours, and Boston Dynamics is aiming to release it commercially in 2022.

Deepfake "Amazon workers" are sowing confusion on Twitter. That's not the problem.

A group of new Twitter accounts, purportedly belonging to Amazon warehouse workers, published many obviously humorous tweet to defend Amazon ahead of a vote on forming labor unions in US-based Amazon warehouses. These accounts all used deepfake profile photos, and they were later revealed to have been created as parody accounts, not by Amazon as a part of a PR campaign. While this particular application of deepfakse seem harmless enough, it highlights the potential uses of deepfakes in misinformation campaigns, especially as the technology improves and deepfakes become harder and harder to spot. Intead of trying to tell whether a photo is real or not, experts suggest looking beyond the profile picture itself for verifying the authenticity of social media accounts:

The giveaway is not that you’ve somehow interrogated the image. It’s that you look at the account and it was created a week ago, or it’s a journalist who claims to be a journalist, but they’ve never written anything else that you could find on a Google search.

Podcast

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

News

Advances & Business

Ambi Robotics, formerly Ambidextrous, raises $6.1M for picking robots - "Ambi Robotics, formerly known as Ambidextrous Laboratories, raised $6.1 million in seed funding for its picking robots and operating system that are based on simulation-to-reality artificial intelligence (AI)."
AI technology detects "ticking time bomb" arteries - "System which identifies patients at risk of fatal heart attack obtains CE mark. The EU has approved an artificial intelligence (AI) technology that can identify people at risk of a fatal heart attack, years before it strikes."
Alphabet Waymo self-driving unit CEO stepping down - "The head of Alphabet Inc's self-driving unit Waymo said on Friday he was stepping down as chief executive and would be replaced by a pair of executives."

Concerns & Hype

U.S. safety agency reviewing 23 Tesla crashes; three from recent weeks - "The U.S. auto safety agency disclosed on Thursday it has opened 25 investigations into crashes of Tesla vehicles, including three crashes in recent weeks, and 23 remain active."
MIT study finds "systematic" labeling errors in popular AI benchmark datasets - "A new paper and website published by researchers at MIT instill little confidence that popular test sets in machine learning are immune to labeling errors."

Expert Opinions & Discussion within the field

Silicon Valley leaders think A.I. will one day fund free cash handouts. But experts aren't convinced - "Artificial intelligence companies could become so powerful and so wealthy that they're able to provide a universal basic income to every man, woman and child on Earth. In as little as 10 years, AI could generate enough wealth to pay every adult in the U.S."
Stop Calling Everything AI, Machine-Learning Pioneer Says - "Michael I. Jordan explains why today's artificial-intelligence systems aren't actually intelligent"

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

Last Week in AI