Discover more from Last Week in AI

Weekly text and audio summaries of the most interesting AI news, as well as editorials commenting on recent events

Over 23,000 subscribers

News

Last Week in AI #142: Tesla FSD crash, French AI strategy, CMU opens AI maker space, and more!

Tesla car using FSD crashes in LA with no injuries but severe car damage, new French AI strategy puts innovation before regulation, CMU's new AI maker space allows students to experiment with robots

Last Week in AI

Nov 15, 2021

Tesla vehicle in ‘Full Self-Driving’ beta mode ‘severely damaged’ after crash in California

A Tesla car using the company’s beta “Full Self-Driving” (FSD) software has crashed near Los Angeles. Luckily no one was injured, but the car was “severely damaged.” According to the report written by the Tesla driver and submitted the National Highway Traffic Safety Administration:

The Vehicle was in FSD Beta mode and while taking a left turn the car went into the wrong lane and I was hit by another driver in the lane next to my lane. the car gave an alert 1/2 way through the turn so I tried to turn the wheel to avoid it from going into the wrong lane but the car by itself took control and forced itself into the incorrect lane creating an unsafe maneuver putting everyone involved at risk.

Tesla has recently allowed drivers to opt in to use the ‘beta’ versions of FSD, something that's not tested enough to roll out for all Tesla owners. Some bugs have been found, and beta software updates have been retracted. This appears to be the the first crash that has resulted from this beta program, and it is likely to put Tesla under more scrutiny by regulators.

Our take: The idea to let non-Tesla employees ‘beta test’ self driving software on public roads is quite strange. Beta testing (“a pre-release of software that is given out to a large group of users to try under real conditions”) is meant to help developers iron out final bugs before release and is a common practice. But unlike using beta versions of software, doing so for cars immediately puts both the drivers and those around them in danger. Should there even be beta testing for autonomous cars by end users? Probably not, or at least not until it's well regulated.

French AI strategy: Tech sector to receive over €2 bln in next 5 years

France has begun to lay out its plan for competing in the AI race as well. On November 8, government ministers presented the second phase of a national AI strategy. More than 2 billion euros (1.5 billion from public funding and 506 million from private) will be poured into French technology in order to produce national and European leaders. In addition to focusing funding on "embedded, frugal, and trusted AI," the nation aims to increase its talent pool by training and financing more students at the undergraduate and graduate levels.

Interestingly, France has also taken a different stance from the EU more broadly on the regulation vs innovation debate: Digital Secretary of State Cédric O lamented the European trend of "regulation before innovation" that could put a severe damper on progress. While regulation might be important, French leaders see adopting regulation a priori before seeing the abilities and limitations of AI technologies to be a futile endeavor. Whatever regulation France may adopt, its leaders would like to leave room for experimentation.

CMU opens first AI maker space to let students 'sharpen the cutting edge of AI'

Carnegie Mellon University has recently opened a maker space dedicated to AI for students to explore and experiment in. While the notion of a university maker space is common - a large room, usually with equipment for rapid hardware prototyping (e.g. CNC machines and 3D printers), shared among students - this AI maker space is the first of its kind. The room has a drone net, the Fetch mobile robot, and other types of robots that students can program and with which to test out the latest AI algorithms.

Our take: This development is in line with recent trends where AI and robotics development are becoming more affordable and accessible. In a few years time we may see universities "loaning" out such hardware like books in a library, or even see similar AI maker spaces at the high school level.

Other News

Research

Reinforcement learning makes for shitty AI teammates in co-op games - "An MIT study finds reinforcement learning frustrates humans in teamplay — here's what that spells for this paradigm of ML in other areas"

AI skin cancer diagnoses risk being less accurate for dark skin – study - "AI systems being developed to diagnose skin cancer run the risk of being less accurate for people with dark skin, research suggests."

DeepMind’s One Pass ImageNet: A New Benchmark for Resource Efficiency in Deep Learning - "Launched at CVPR 12 years ago, the ImageNet database now comprises more than 14 million labelled images and has become one of the most influential benchmarks in the field of computer vision."

University of Waterloo AI Researchers Introduce A New NLP Model ‘AfriBERTa’ For African Languages Using Deep Learning Techniques - "A technology that has been around for years but most often taken for granted is Natural Language Processing(NLP). It is the employment of computational methods to analyze and synthesize natural language and speech."

Applications

Nvidia wants to fill the virtual and physical worlds with AI avatars - "Nvidia has announced a new platform for creating virtual agents named Omniverse Avatar."

NVIDIA's new AI brain for robots is six times more powerful than its predecessor - "NVIDIA has launched a follow-up to the Jetson AGX Xavier, its $1,100 AI brain for robots that it released back in 2018. The new module, called the Jetson AGX Orin, has six times the processing power of Xavier even though it has the same form factor and can still fit in the palm of one's hand."

Nvidia’s Riva Custom Voice lets companies create custom voices powered by AI - "At its fall 2021 GPU Technology Conference (GTC), Nvidia unveiled Riva Custom Voice, a new toolkit that the company claims can enable customers to create custom, “human-like” voices with only 30 minutes of speech recording data."

AI customer service bots are making their way to a store near you - "You won't be able to tell if it's a real human or an AI chatbot."

Why Many People Are Turning Toward Robots And AI To Help Support Their Mental Health And Careers - "The pandemic was more than a devastating health crisis. The ripple effects changed the way we work and live our lives. While this may seem insensitive, the Covid-19 outbreak led to some positive changes to our culture."

Mozrt, a Deep Learning Recommendation System Empowering Walmart Store Associates with a… - "Walmart employs nearly 1.6 million associates staffed at over 4,700 stores in the U.S. Each associate is responsible for completing various tasks, which are often updated and revised based on their daily schedule and assignment."

Business

Walmart is using fully driverless trucks to ramp up its online grocery business - "Walmart and Silicon Valley start-up Gatik said that, since August, they’ve operated two autonomous box trucks, without a safety driver, on a 7-mile loop daily for 12 hours."

Philips augments cardiac AI diagnostic programs with Cardiologs acquisition - "Following a multibillion-dollar acquisition late last year to grow its cardiac connected care franchise, Philips is continuing the effort with a new deal to take over Cardiologs, a French startup developing artificial-intelligence-powered diagnostics."

MLOps platform Landing AI raises $57M to help manufacturers adopt computer vision - "Palo Alto, California-based Landing AI, the AI startup led by Andrew Ng — the cofounder of Google Brain, one of Google’s AI research divisions — today announced that it raised $57 million in a series A funding round led by McRock Capital. "

AI-driven search engine You.com takes on Google with $20M - "You.com, which bills itself as the world’s first open source search engine, today announced its public beta launch along with $20 million in funding led by Salesforce CEO Marc Benioff with participation from Breyer Capital, Sound Ventures, Day One Ventures, and others."

Inworld AI joins metaverse innovation with AI-driven virtual characters - "Inworld AI, a company that aims to develop a platform that enables users to create AI-driven virtual characters which can be used to populate virtual worlds, announced today that it has raised $7 million in seed funding."

Self-driving truck startup Kodiak Robotics raises $125M - "Kodiak Robotics, one of the last privately held autonomous truck outfits in Silicon Valley, has raised $125 million in new capital, funding that will be used to double its workforce as it pursues a path to commercialization."

Cerebras Completes Series F Funding, Another $250M for $4B Valuation - "Every once in a while, a startup comes along with something out of left field. In the AI hardware generation, Cerebras holds that title, with their Wafer Scale Engine. The second generation product, built on TSMC 7nm, is a full wafer packed to the brim with cores, memory, and performance."

Report: AI startup funding hits record high of $17.9B in Q3 - "Even as economies struggle with the chaos of the pandemic, the AI startup space continues to grow stronger with increased investments and M&A deals. According to the latest State of AI report from CB Insights, the global funding in the segment has seen a significant surge, growing from $16."

Companies Order Record Number of Robots Amid Labor Shortage - "North American firms ordered nearly 29,000 robotics units in first nine months of 2021Resume Subscription"

Concerns

Current AI Practices Could Be Enabling a New Generation of Copyright Trolls - "A new research collaboration between Huawei and academia suggests that a great deal of the most important current research in artificial intelligence and machine learning could be exposed to litigation as soon as it becomes commercially prominent, because the datasets that make breakthroughs possibl"

EU: Artificial Intelligence Regulation Threatens Social Safety Net - "The European Union’s plan to regulate artificial intelligence is ill-equipped to protect people from flawed algorithms that deprive them of lifesaving benefits and discriminate against vulnerable populations, Human Rights Watch said in report on the regulation released today."

Timnit Gebru Says Artificial Intelligence Needs to Slow Down - "Artificial intelligence researchers are facing a problem of accountability: How do you try to ensure decisions are responsible when the decision maker is not a responsible person, but rather an algorithm? Right now, only a handful of people and organizations have the power—and resources—to autom"

Algorithmic tracking is ‘damaging mental health’ of UK workers - "Monitoring of workers and setting performance targets through algorithms is damaging employees’ mental health and needs to be controlled by new legislation, according to a group of MPs and peers."

Bias in A.I. is a big, thorny, ethical issue - "A.I. data are typically drawn from biased— even racist— data sets. And ethicists say it's time to confront the issue."

Israel escalates surveillance of Palestinians with facial recognition program in West Bank - "The Israeli military has been conducting a broad surveillance effort in the occupied West Bank to monitor Palestinians by integrating facial recognition with a growing network of cameras and smartphones, according to descriptions of the program by recent Israeli soldiers."

Analysis

Rise of the Robots Speeds Up in Pandemic With U.S. Labor Scarce - "American workers are hoping that the tight pandemic labor market will translate into better pay. It might just mean robots take their jobs instead. Labor shortages and rising wages are pushing U.S. business to invest in automation."

Can and should we trust AI? - "You have probably heard that in April 2021, the European Commission put forward a proposal to regulate Artificial Intelligence (AI) so that the European Union can reap the benefits of AI. The words trust and trustworthiness together appear over 100 times in their proposal."

Policy

Proposed Law to Require Social Media to Offer Algorithm-Free Version - "A bipartisan group of House lawmakers has introduced a new bill that would make it mandatory for services that use algorithms to serve content to offer a version that allows users to turn that feature off."