A 40cm-square patch that renders you invisible to person-detecting AIs

Researchers from KU Leuven have published a paper showing how they can create a 40cm x 40cm "patch" that fools a convoluted neural network classifier that is otherwise a good tool for identifying humans into thinking that a person is not a person -- something that could be used to defeat AI-based security camera systems. They theorize that the could just print the patch on a t-shirt and get the same result. Read the rest

A machine-learning wishlist for hardware designers

Pete Warden (previously) is one of my favorite commentators on machine learning and computer science; yesterday he gave a keynote at the IEEE Custom Integrated Circuits Conference, on the ways that hardware specialization could improve machine learning: his main point is that though there's a wealth of hardware specialized for creating models, we need more hardware optimized for running models. Read the rest

Ford CEO: we "overestimated" self-driving cars

Ford CEO Jim Hackett -- formerly head of the company's autonomous vehicle division -- publicly announced that the company had "overestimated the arrival of autonomous vehicles" and that the vehicles, when they did arrive, their "applications will be narrow, what we call geo-fenced, because the problem is so complex." Read the rest

Artist designs a machine-learning assisted sculpture, then casts it in the powdered remains of the computer used to design it

Ben Snell's sculpture Dio was created by training a machine learning system on a corpus of 1,000+ sculptures, tweaked in some unspecified way by Snell, who then 3D printed a mold based on the final shape: he filled the mold with a resin impregnated with the computer that ran the algorithm, which Snell had ground to powder. Read the rest

Amazon stores recordings of Alexa interactions and turns them over to internal staff and outside contractors for review

Bloomberg reporters learned that -- despite public pronouncements to the contrary -- Amazon has an "annotation team" of thousands of people all over the world, charged with reviewing recordings made by Alexa devices in the field, with both staffers and contractors listening to conversations that Alexa owners have had with and near their devices. Read the rest

What the rest of the world doesn't know about Chinese AI

ChinAI Jeff Ding's weekly newsletter reporting on the Chinese AI scene; on the occasion of the newsletter's first anniversary, Ding has posted a roundup of things about the Chinese AI scene that the rest of the world doesn't know about, or harbors incorrect beliefs about. Read the rest

Googler uprising leads to shut down of AI ethics committee that included the president of the Heritage Foundation

This week, thousands of googlers and many others (including me) signed an open letter objecting to the inclusion of Heritage Foundation president Kay Coles James on the company's Advanced Technology External Advisory Council (ATEAC), on the the grounds that James had frequently evinced viciously transphobic, racist, anti-immigrant sentiments. Read the rest

Small stickers on the ground trick Tesla autopilot into steering into opposing traffic lane

Researchers from Tencent Keen Security Lab have published a report detailing their successful attacks on Tesla firmware, including remote control over the steering, and an adversarial example attack on the autopilot that confuses the car into driving into the oncoming traffic lane. Read the rest

The Chinese Communist Party's newspaper has spun out an incredibly lucrative censorship business

People.cn is a publicly listed subsidiary of The People's Daily, the official newspaper of the Chinese Communist Party; its fortunes are rising and rising with no end in sight as it markets itself as an outsource censorship provider who combine AI and a vast army of human censors to detect and block attempts to circumvent censorship through irony, memes, and metaphors. Read the rest

Philadelphia city council candidate says his secret AI has discovered disqualifying fraud in the nominations of 30 out of 33 candidates

Devon Cade -- a former bureaucrat who now describes himself as a "philanthropist" -- has asked a court to disqualify 30 out of the 33 other Democrats standing in the primary for the city's council elections on the grounds that the signatures on their nominating petitions were forged. Read the rest

Some pretty impressive machine-learning generated poetry courtesy of GPT-2

GPT-2 is Open AI's language-generation model (last seen around these parts as a means of detecting machine-generated text); it's powerful and cool, and Gwern Branwen fed it the Project Gutenberg poetry corpus to see what kind of poetry it would write. Read the rest

Palmer Luckey wins secretive Pentagon contract to develop AI for drones

Palmer Luckey (previously) the alt-right financier who was made a billionaire by Mark Zuckerberg's decision to acquire his VR startup Oculus, is now running a Peter-Thiel-backed surveillance startup called Anduril Industries, which has won a contract to contribute to Project Maven, the Pentagon's controversial AI-for-drones system (Google's involvement in Project Maven sparked an employee uprising that ended with the relevant executives leaving the company and the contract being allowed to lapse). Read the rest

A machine-learning system that guesses whether text was produced by machine-learning systems

Gltr is an MIT-IBM Watson Lab/Harvard NLP joint project that analyzes texts and predicts whether that text was generated by a machine-learning model. Read the rest

Towards a general theory of "adversarial examples," the bizarre, hallucinatory motes in machine learning's all-seeing eye

For several years, I've been covering the bizarre phenomenon of "adversarial examples (AKA "adversarial preturbations"), these being often tiny changes to data than can cause machine-learning classifiers to totally misfire: imperceptible squeaks that make speech-to-text systems hallucinate phantom voices; or tiny shifts to a 3D image of a helicopter that makes image-classifiers hallucinate a rifle Read the rest

Common Voice: Mozilla releases the largest dataset of voice samples for free, for all

42,000 Mozilla supporters contributed to Common Voice, a free-open dataset of 1,361 hours of voice recordings in 18 languages, which is now free for anyone to use as a set of "high quality, transcribed voice data... available to startups, researchers, and anyone interested in voice-enabled technologies" -- in a field plagued with sampling bias problems, this is a dataset that aims to be diverse, representative and inclusive, and it's growing by the day (you can contribute your voice too!) -- the whole project is inspiring. (via Four Short Links) Read the rest

Beyond GIGO: how "predictive policing" launders racism, corruption and bias to make them seem empirical

"Predictive policing" is the idea that you can feed crime stats to a machine-learning system and it will produce a model that can predict crime. It is garbage. Read the rest

Chasing down that list of potential Predpol customers reveals dozens of cities that have secretly experimented with "predictive policing"

Last October, I published a list of cities that appeared to have contracted with Predpol, a "predictive policing" company, based on research provided to me by an anonymous source who used clever methods to uncover the customer list. Read the rest

More posts