Archives 4 February 2023

A Bot Was Scheduled To Argue In Court, Then Came the Jail Threats

A British man who planned to have a “robot lawyer” help a defendant fight a traffic ticket has dropped the effort after receiving threats of possible prosecution and jail time. […] The first-ever AI-powered legal defense was set to take place in California on Feb. 22, but not anymore. As word got out, an uneasy buzz began to swirl among various state bar officials, according to Browder. He says angry letters began to pour in. “Multiple state bar associations have threatened us,” Browder said. “One even said a referral to the district attorney’s office and prosecution and prison time would be possible.” In particular, Browder said one state bar official noted that the unauthorized practice of law is a misdemeanor in some states punishable up to six months in county jail.

“Even if it wouldn’t happen, the threat of criminal charges was enough to give it up,” [said Joshua Browden, the CEO of the New York-based startup DoNotPay]. “The letters have become so frequent that we thought it was just a distraction and that we should move on.” State bar associations license and regulate attorneys, as a way to ensure people hire lawyers who understand the law. Browder refused to cite which state bar associations in particular sent letters, and what official made the threat of possible prosecution, saying his startup, DoNotPay, is under investigation by multiple state bar associations, including California’s.

Seattle schools sue TikTok, Meta and other platforms over youth ‘mental health crisis’

Seattle public schools have sued the tech giants behind TikTok, Facebook, Instagram, YouTube and Snapchat, accusing them of creating a “mental health crisis among America’s Youth.” The 91-page lawsuit filed in a US district court states that tech giants exploit the addictive nature of social media, leading to rising anxiety, depression and thoughts of self-harm.

“Defendants’ growth is a product of choices they made to design and operate their platforms in ways that exploit the psychology and neurophysiology of their users into spending more and more time on their platforms,” the complaint states. “[They] have successfully exploited the vulnerable brains of youth, hooking tens of millions of students across the country into positive feedback loops of excessive use and abuse of Defendants’ social media platforms.”

Harmful content pushed to users includes extreme diet plants, encouragement of self-harm and more, according to the complaint. That has led to a 30 percent increase between 2009 and 2019 of students who report feeling “so sad or hopeless… for two weeks or more in a row that [they] stopped doing some usual activities.”

Boston Dynamics’ latest Atlas video demos a robot that can run, jump and now grab and throw

Boston Dynamics released a demo of its humanoid robot Atlas, showing it pick up and deliver a bag of tools to a construction worker. While Atlas could already run and jump over complex terrain, the new hands, or rudimentary grippers, “give the robot new life,” reports TechCrunch. From the report:
The claw-like gripper consists of one fixed finger and one moving finger. Boston Dynamics says the grippers were designed for heavy lifting tasks and were first demonstrated in a Super Bowl commercial where Atlas held a keg over its head. The videos released today show the grippers picking up construction lumber and a nylon tool bag. Next, the Atlas picks up a 2×8 and places it between two boxes to form a bridge. The Atlas then picks up a bag of tools and dashes over the bridge and through construction scaffolding. But the tool bag needs to go to the second level of the structure — something Atlas apparently realized and quickly throws the bag a considerable distance. Boston Dynamics describes this final maneuver: ‘Atlas’ concluding move, an inverted 540-degree, multi-axis flip, adds asymmetry to the robot’s movement, making it a much more difficult skill than previously performed parkour.”

Researchers Are Getting Eerily Good at Using WiFi to ‘See’ People Through Walls in Detail

Researchers at Carnegie Mellon University developed a method for detecting the three dimensional shape and movements of human bodies in a room, using only WiFi routers. From a report:
To do this, they used DensePose, a system for mapping all of the pixels on the surface of a human body in a photo. DensePose was developed by London-based researchers and Facebook’s AI researchers. From there, according to their recently-uploaded preprint paper published on arXiv, they developed a deep neural network that maps WiFi signals’ phase and amplitude sent and received by routers to coordinates on human bodies. Researchers have been working on “seeing” people without using cameras or expensive LiDAR hardware for years. In 2013, a team of researchers at MIT found a way to use cell phone signals to see through walls; in 2018, another MIT team used WiFi to detect people in another room and translate their movements to walking stick-figures.

Meta Sues Surveillance Company for Scraping Data With Fake Facebook Accounts

Meta has filed a legal complaint against a company for allegedly creating tens of thousands of fake Facebook accounts to scrape user data and provide surveillance services for clients. From a report:
The firm, Voyager Labs, bills itself as “a world leader in advanced AI-based investigation solutions.” What this means in practice is analyzing social media posts en masse in order to make claims about individuals. In 2021, for example, The Guardian reported how Voyager Labs sold its services to the Los Angeles Police Department, with the company claiming to predict which individuals were likely to commit crimes in the future.

Meta announced the legal action in a blog post on January 12th, claiming that Voyager Labs violated its terms of service. According to a legal filing issued on November 11th, Meta alleges that Voyager Labs created over 38,000 fake Facebook user accounts and used its surveillance software to gather data from Facebook and Instagram without authorization. Voyager Labs also collected data from sites including Twitter, YouTube, and Telegram.

Microsoft’s new AI can simulate anyone’s voice with 3 seconds of audio

Microsoft researchers announced a new text-to-speech AI model called VALL-E that can closely simulate a person’s voice when given a three-second audio sample. Once it learns a specific voice, VALL-E can synthesize audio of that person saying anything — and do it in a way that attempts to preserve the speaker’s emotional tone. Its creators speculate that VALL-E could be used for high-quality text-to-speech applications, speech editing where a recording of a person could be edited and changed from a text transcript (making them say something they originally didn’t), and audio content creation when combined with other generative AI models like GPT-3.

Microsoft calls VALL-E a “neural codec language model,” and it builds off of a technology called EnCodec, which Meta announced in October 2022. Unlike other text-to-speech methods that typically synthesize speech by manipulating waveforms, VALL-E generates discrete audio codec codes from text and acoustic prompts. It basically analyzes how a person sounds, breaks that information into discrete components (called “tokens”) thanks to EnCodec, and uses training data to match what it “knows” about how that voice would sound if it spoke other phrases outside of the three-second sample. Or, as Microsoft puts it in the VALL-E paper (PDF): “To synthesize personalized speech (e.g., zero-shot TTS), VALL-E generates the corresponding acoustic tokens conditioned on the acoustic tokens of the 3-second enrolled recording and the phoneme prompt, which constrain the speaker and content information respectively. Finally, the generated acoustic tokens are used to synthesize the final waveform with the corresponding neural codec decoder.”

[…] While using VALL-E to generate those results, the researchers only fed the three-second “Speaker Prompt” sample and a text string (what they wanted the voice to say) into VALL-E. So compare the “Ground Truth” sample to the “VALL-E” sample. In some cases, the two samples are very close. Some VALL-E results seem computer-generated, but others could potentially be mistaken for a human’s speech, which is the goal of the model. In addition to preserving a speaker’s vocal timbre and emotional tone, VALL-E can also imitate the “acoustic environment” of the sample audio. For example, if the sample came from a telephone call, the audio output will simulate the acoustic and frequency properties of a telephone call in its synthesized output (that’s a fancy way of saying it will sound like a telephone call, too). And Microsoft’s samples (in the “Synthesis of Diversity” section) demonstrate that VALL-E can generate variations in voice tone by changing the random seed used in the generation process.

Microsoft has not provided VALL-E code for others to experiment with, likely to avoid fueling misinformation and deception.

‘Please Don’t Film Me in 2023’

Many viewers on TikTok ate it up, but others pushed back on the idea that there’s humor in filming and posting an unsuspecting neighbor for content. This year, I saw more and more resistance to the practice that’s become normal or even expected…. [P]eople who have been featured in videos unbeknownst to them have pointed out that even if there’s no ill will, it’s just unnerving and weird to be filmed by others as if you’re bit characters in the story of their life. One TikTok user, @hilmaafklint, landed in a stranger’s vlog when they filmed her to show her outfit. She didn’t realize it had happened until another stranger recognized her and tagged her in the video.

“It’s weird at best, and creepy and a safety hazard at worst,” she says in a video….

Even before TikTok, public space had become an arena for constant content creation; if you step outside, there’s a chance you’ll end up in someone’s video. It could be minimally invasive, sure, but it could also shine an unwanted spotlight on the banal moments that just happen to get caught on film. This makeshift, individualized surveillance apparatus exists beyond the state-sponsored systems — the ones where tech companies will hand over electronic doorbell footage without a warrant or where elected officials allow police to watch surveillance footage in real time. We’re watched enough as it is.

So if you’re someone who makes content for the internet, consider this heartfelt advice and a heads-up. If you’re filming someone for a video, please ask for their consent.

And if I catch you recording me for content, I will smack your phone away.