Last Week in AI #130: Tesla's new bot, 'foundation' models, the poetry of AI Art

Tesla unveiled plans for a humanoid, 'Tesla bot,' Stanford researchers discuss opportunities and risks of large AI models, how misunderstandings of AI art can be poetic, and more!

Aug 23, 2021

Elon Musk unveils 'Tesla bot,' a humanoid robot that would be made from Tesla's self-driving AI

On August 18th, Tesla held an “AI Day”, similar to the “Tesla Autonomy Day” it held in 2019 and the “Tesla Battery Day” in 2020. It was stated by Elon Musk that “convincing the best AI talent to join Tesla is the sole goal.” The majority of the presentation had to deal with the technical approach Tesla researchers and engineers are taking to improve their cars’ self driving capabilities, both in terms of software and hardware.

However, most headlines about AI Day had to do with a short segment near the end when Musk announced that Tesla was working on a "Tesla Bot," a humanoid robot capable of doing various tasks that humans can. Some specifications were presented, but it is unclear how far along the design of the robot is; Musk stated Tesla may unveil a prototype of it in 2022. The general response has been one of excitement and interest, though many also responded skeptically given the difficulty of the task and Musk’s history of making predictions.

100+ Standford Researchers Publish 200+ Page Paper on the AI Paradigm Shift Introduced by Large-Scale Models

ai - artificial intelligence and deep learning concept of neural networks. Wave equalizer. Blue and purple lines. Vector illustration

Last week a large group of AI researchers from Stanford published a review paper on what they call foundation models, the type of large-scale machine learning models popularized over the last two years, like OpenAI's GPT-3. The study gives an in-depth discussion of future opportunities as well as risks of scaling up such models. In particular, it seems like there is a trend of homogenoization in AI applications, where many fields, from speech and vision to healthcare and robotics, are adopting the use of large models. Such homogenoization is beneficial for domains where task-specific data is limited. However, it also amplifies the inherent risks of such approaches, like bias and poor explainability. The Stanford researchers also started a workshop on foundation models to advance both the development and deployment of these models in the future.

Appreciating the Poetic Misunderstandings of A.I. Art

Not long ago, the AI art scene exploded as people found ways to hook up OpenAI's CLIP with generative adversarial networks. As twenty-three-year-old artist Ryan Murdock discovered, you can "reverse" CLIP's process by giving it text like "temple" and iteratively updating an image until CLIP determined it looked like a temple. Sam Burton-King, a twenty-year-old student at Northwestern University, used this same technology as the basis for @images_ai, a Twitter account promising to create AI-generated art of a user's request. While the images that pop out are interesting, they are inhuman: Burton-King commented on the fact that these images tend to include tiles and are incoherent as a whole.

Other News

Research

Deep machine learning study finds that body shape is associated with income - "A new study published in PLOS One has found a relationship between a person’s body shape and their family income."

Deepmind Introduces PonderNet, A New AI Algorithm That Allows Artificial Neural Networks To Learn To “Think For A While” Before Answering - "Deepmind introduces PonderNet, a new algorithm that allows artificial neural networks to learn to think for a while before answering."

Boston Dynamics' robots can parkour better than you - "Don't expect an easy getaway if one of Boston Dynamics' Atlas robots ever chases you down. The Hyundai-owned firm has shared a video (below) of the humanoid bots successfully completing a parkour routine in an obstacle course for the first time."

Supporting COVID-19 policy response with large-scale mobility-based modeling - "Mobility restrictions, from stay-at-home orders to indoor occupancy caps, have been utilized extensively by policymakers during the COVID-19 pandemic. These reductions in mobility help to control the spread of the virus 12, but they come at a heavy cost to businesses and employees."

Updates and Lessons from AI Forecasting - "Earlier this year, my research group commissioned 6 questions for professional forecasters to predict about AI. Broadly speaking, 2 were on geopolitical aspects of AI and 4 were on future capabilities: Forecasters output a probability distribution over outcomes for 2022, 2023, 2024, and 2025."

AI Models Aid in Predicting Risk of Breast, Lung Cancer - "Along with aiding radiologists in detecting breast cancer, AI can help in a number of other applications, including measuring breast density – a known risk factor for cancer. In a 2018 Radiology study, “Mammographic Breast Density Assessment Using Deep Learning: Clinical Implementation,” Drs."

This AI Helps Detect Wildlife Health Issues in Real Time - "During the spring, a troublesome pattern plays out as marine birds along the California coast die from domoic acid poisoning, which is caused by harmful algal blooms."

Applications

Using Artificial Intelligence to Sniff Out Corporate Greenwashers - "GreenWatch is a new tool for investors worried that some sustainability claims are too good to be true."

Using Deep Learning Algorithms to Give Bicyclists the ‘Green Wave’ at Traffic Signals - "Led by Dr. Stephen Fickas of the University of Oregon (UO), transportation researchers are working to give bicyclists smoother rides by allowing them to communicate with traffic signals via a mobile app."

Face ID could help protect Afghan allies against US face recognition kit now in Taliban hands - "One of the many worrying developments in the tragedy unfolding in Afghanistan is that US face recognition and fingerprint devices and databases have been seized by the Taliban."

AI gave Val Kilmer his voice back. But critics worry the technology could be misused. - "Val Kilmer lost his natural speaking voice in 2015 during a battle with throat cancer. The Hollywood actor worked with an AI firm to digitally restore it."

Business

Samsung Has Its Own AI-Designed Chip. Soon, Others Will Too - "Samsung is using artificial intelligence to automate the insanely complex and subtle process of designing cutting-edge computer chips. The South Korean giant is one of the first chipmakers to use AI to create its chips."

AI Firm SenseTime Said to Tap HSBC For $2 Billion Hong Kong IPO - "SenseTime Group Ltd., China’s largest artificial intelligence company, is working with HSBC Holdings Plc to arrange its planned Hong Kong initial public offering that could raise at least $2 billion, according to people familiar with the matter. China International Capital Corp."

Waymo is building a hub for its autonomous trucks in Texas - "Waymo announced plans to build a hub for its autonomous semi-trailer trucks on a nine-acre site near Dallas-Fort Worth, Texas."

China's Baidu launches second chip and a 'robocar' as it sets up future in AI and autonomous driving - "GUANGZHOU, China — Chinese internet giant Baidu unveiled its second-generation artificial intelligence chip, its first "robocar" and a rebranded driverless taxi app, underscoring how these new areas of technology are key to the company's future growth."

Startup That Develops AI-Pumped Robotic Arms Raises $37 Million - "Rapid Robotics’ Series B was led by Kleiner Perkins and Tiger Global, and comes as labor and parts shortages are hitting the economy"

Concerns

Why Teslas Keep Striking Parked Firetrucks and Police Cars - "On Monday, the National Highway Traffic Safety Administration opened an investigation into Tesla."

‘It Happened So Fast’: Inside a Fatal Tesla Autopilot Accident - "George Brian McGee, a finance executive in Florida, was driving home in a Tesla Model S operating on Autopilot, a system that can steer, brake and accelerate a car on its own, when he dropped his phone during a call and bent down to look for it."

Reddit user reverse-engineers what they believe is Apple's solution to flag child sex abuse - "Earlier this month, Apple announced that it is introducing new child safety features for its entire ecosystem. As a part of this effort, the Cupertino firm is going to scan the contents of iCloud and the Messages app using on-device machine learning to detect Child Sexual Abuse Material (CSAM)."

How AI-powered Tech Landed Man In Jail With Scant Evidence - "Michael Williams’ wife pleaded with him to remember their fishing trips with the grandchildren, how he used to braid her hair, anything to jar him back to his world outside the concrete walls of Cook County Jail."

‘A lot of people are sleepwalking into it’: the expert raising concerns over AI - "It’s one of the most profound innovations of our time - and Manhattan-based Australian Kate Crawford wants us to wake up to AI’s inherent risks. Kate Crawford, one of the world’s pre-eminent scholars on the social and political implications of artificial intelligence, is being watched."

Using 'AI-based software like Proctorio and ProctorU' to monitor online exams is a really bad idea, says uni panel - "A committee at the University of Texas in Austin has advised against using AI software to oversee students' online tests, citing the psychological toll on students and the financial toll on academic institutions."

Analysis

Philip Glass on Artificial Intelligence and Art - "This conversation with the composer Philip Glass and me discusses an exciting project in partnership with OpenAi, in which we trained a neural net on a corpus of Glass’ work. He offers commentary on the music created by “his AI”, as well as insights on composition and creating art."

What It’s Like to Write Articles with Artificial Intelligence - "I see GPT-3 both as a threat to the conventional notion of writing, but also as a great new tool for authors. Perhaps an analogy is useful here: I’m a calligrapher, and I learn about the Gutenberg press."

Policy

U.S. opens formal safety probe into some 765,000 Tesla vehicles - "WASHINGTON, Aug 16 (Reuters) - U.S. auto safety regulators said early Monday they have opened a formal safety probe into Tesla Inc's driver assistance system Autopilot after a series of crashes involving emergency vehicles."

Senators question Justice Department funding for AI-powered policing tech - "CHICAGO — A Democratic senator said the U.S. Justice Department needs to look into whether the algorithm-powered police technologies it funds contribute to racial bias in law enforcement and lead to wrongful arrests. Sen."

Podcast

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

Last Week in AI