AI Weekly: Charlatan AI is a public nuisance

In a hilarious turn of events last month, a Russian robot named Boris was unmasked as a man in a robot suit.

Likewise, state-run media in China unveiled its AI reporter in November, and to this day it’s not clear if this is an actual AI system boiling down news stories or just a synthesized voice with an avatar.

More fabricated robotic theatrics appeared to be on display this week at the Consumer Electronics Show in Las Vegas, where LG CTO I.P. Park delivered the opening CES keynote address.

Park was joined onstage for the hour-long presentation by CLOi, a conceptual robot perhaps best known for failing during a live demo at CES a year ago.

This year, however, CLOi did a bit of everything: The robot acted as cohost, cracked jokes, delivered some LG HomeBrew beer, and even helped some guy who hates blind dates find true love.

Its prowess with love — and, more importantly, its conversational AI — seemed too good to be true. The things CLOi said onstage did not appear to be a demonstration of what the robot’s NLP can actually do, but rather a script that the robot read to the audience, and that’s a disservice.

LG did not respond to repeated questions about whether the responses the robot shared onstage were scripted or in any way generated by an AI system, but here’s a clip of the presentation.

Whatever CLOi is actually able to do, the performance by the robot brought to mind Facebook’s AI chief calling Hanson Robotics’ Sophia bullshit a year ago, as well as a conversation I had last year with a business executive who said he was being forced to talk about Duplex and address misguided questions about what’s possible today rather than talk about his product.

That’s why tech companies should spare the world overblown or fabricated pitches of what their AI can do.

False perceptions about what is and isn’t possible can negatively impact public perception, business decisions, and even government policy. More lawmakers talked about AI last year in Congress and the U.K.’s Parliament, according to the AI Index report. Sharing her predictions for the year ahead with VentureBeat , Accenture’s responsible AI lead Rumman Chowdhury said she expects to see more regulation of AI in the year ahead.

False perceptions can also take oxygen away from more important issues surrounding AI, such as bias or the concerns of people worried their job will disappear.

One of the toughest things about artificial intelligence today is the extent to which it makes humans question what’s real and what’s fake. Yes, deepfakes and artificial intelligence that can manipulate images have a lot to do with that, but it’s also the result of overzealous marketing and overblown claims of what they can do.

When artificial intelligence is wrapped up in tremendous potential and an equal amount of misleading hype, the average person could be forgiven for sometimes being duped by high-profile marketing ploys, but the confusion this creates can have real consequences.

Fabricated examples of what an AI system can do take advantage of the hopes and fears of a public being told on a loop that robots will make their lives better but can also take their jobs and become killing machines.

Even if it sometimes means tripping over your own shoelaces and making mistakes onstage, I think demonstrations should present robots with their limitations in order to set realistic expectations.

To be clear: CLOi isn’t a total fabrication: The line of robots is being used to help people in stores and in an airport in South Korea, and LG is investing tens of millions of dollars in its robots. But companies should stick to sharing what’s actually possible, otherwise they’re no better than Boris.

For AI coverage, send news tips to Kyle Wiggers and Khari Johnson — and be sure to bookmark our AI Channel.

Thanks for reading,

Khari Johnson
AI Staff Writer

P.S. Please enjoy this video of the Google Assistant ride at CES, and this writeup on why it’s the kind of thing that CES should be all about by VentureBeat news editor Emil Protalinski.

From VB

Samsung is using computer vision to tell you what’s in your fridge

Artificial intelligence in Samsung’s Family Hub may someday deliver on the perennial CES promise of the fridge of the future, with the introduction of Bixby Vision.

Mobileye plans to deploy fully autonomous cars in 4 years

Mobileye plans to deploy fully autonomous driving tech in the next four years, starting with Israel. We spoke with CEO Amnon Shashua to learn more.

Google Home gets real-time interpretations for 27 languages

Google Home speakers will soon carry out real-time language translations in up to 27 languages, one of several feature upgrades announced today at CES.

Baidu announces Apollo 3.5 and Apollo Enterprise, says it has over 130 partners

Baidu announced a new version of its open source autonomous car platform, Apollo, and launched Apollo Enterprise, a multiservice solution for auto OEMs.

Amazon’s Ring unveils Door View Cam for peephole video and smart outdoor lighting

Amazon unveiled a series of new Ring products, including the Door View Cam for live video from the peephole of your front door, as well as new outdoor lighting and sensors made for motion detection or amplifying security alarms.

Motorola Solutions buys public safety machine-vision specialist VaaS for $445 million

Public safety solutions provider Motorola Solutions has acquired VaaS, maker of AI-driven image capture and analysis technology, for $445 million.

Beyond VB

Venture capitalist: AI will displace 40 percent of world’s jobs in as soon as 15 years

Using a common type of brain scan, researchers programmed a machine-learning algorithm to diagnose early-stage Alzheimer’s disease about six years before a clinical diagnosis is made – potentially giving doctors a chance to intervene with treatment. (via CBS News)

DARPA wants to build an AI to find the patterns hidden in global chaos

That most famous characterization of the complexity causality, a butterfly beating its wings and causing a hurricane on the other side of the world, is thought-provoking but ultimately not helpful. (via TechCrunch)

Meet Caper, the AI self-checkout shopping cart

Caper wants to make eliminating checkout lines as easy as replacing their shopping carts while offering a more familiar experience for customers. (via TechCrunch)