Jonathan Zittrain (previously) is consistently a source of interesting insights that often arrive years ahead of their wider acceptance in tech, law, ethics and culture (2008's The Future of the Internet (and how to stop it) is surprisingly relevant 11 years later); in a new long essay on Medium (shorter version in the New Yorker), Zittrain examines the perils of the "intellectual debt" that we incur when we allow machine learning systems that make predictions whose rationale we don't understand, because without an underlying theory of those predictions, we can't know their limitations. Read the rest

The Scite project has a corpus of millions of scientific articles that it has analyzed with deep learning tools to determine whether any given paper has been supported or contradicted by subsequent publications; you can check Scite via the website, or install a browser plugin version (Firefox, Chrome). (Thanks, Josh!) Read the rest

"Anonymized data" is one of those holy grails, like "healthy ice-cream" or "selectively breakable crypto" -- if "anonymized data" is a thing, then companies can monetize their surveillance dossiers on us by selling them to all comers, without putting us at risk or putting themselves in legal jeopardy (to say nothing of the benefits to science and research of being able to do large-scale data analyses and then publish them along with the underlying data for peer review without posing a risk to the people in the data-set, AKA "release and forget"). Read the rest

Evan Greer from Fight for the Future writes, "Facial recognition might be the most invasive and dangerous form of surveillance tech ever invented. While it's been in the headlines lately, most of us still don't know whether it's happening in our area. My organization Fight for the Future has compiled an interactive map that shows everywhere in the US (that we know of) facial recognition being used -- but also where there are local efforts to ban it, like has already happened in San Francisco, Oakland, and Somerville, MA. We've also got a tool kit for local residents who want to get an ordinance or state legislation passed in their area." Read the rest

In Q2 2018, Chinese investors sank $2.87b into AI startups; in Q2 2019, it was $140.7m. Read the rest

I used to be on the program committee for the O'Reilly Emerging Technology conferences; one year we decided to make the theme "magic" -- all the ways that new technologies were doing things that baffled us and blew us away. Read the rest

In the latest Adafruit video (previously) the proprietors, Limor "ladyada" Friend and Phil Torrone, explain the basics of machine learning, with particular emphasis on the difference between computing a model (hard) and implementing the model (easy and simple enough to run on relatively low-powered hardware), and then they install and run Tensorflow Light on a small, open-source handheld and teach it to distinguish between someone saying "No" and someone saying "Yes," in just a few minutes. It's an interesting demonstration of the theory that machine learning may be most useful in tiny, embedded, offline processors. (via Beyond the Beyond) Read the rest

Joël Franusic became obsessed with Krazy Kat, but was frustrated by the limited availability and high cost of the books anthologizing the strip (some of which were going for $600 or more on Amazon); so he wrote a scraper that would pull down thumbnails from massive archives of pre-1923 newspapers and then identified 100 pages containing Krazy Kat strips to use as training data for a machine-learning model. Read the rest

Rob Smith is an eminent computer scientist and machine learning pioneer whose work on genetic algorithms has been influential in both industry and the academy; now, in his first book for a general audience, Rage Inside the Machine: The Prejudice of Algorithms, and How to Stop the Internet Making Bigots of Us All, Smith expertly draws connections between AI, neoliberalism, human bias, eugenics and far-right populism, and shows how the biases of computer science and the corporate paymasters have distorted our whole society. Read the rest

In The Next Generation of Deep Learning Hardware: Analog Computing *Sci-Hub mirror), a trio of IBM researchers discuss how new materials could allow them to build analog computers that vastly improved the energy/computing efficiency in training machine-learning networks. Read the rest

One of the griftiest corners of late-stage capitalism is the "public safety" industry, in which military contractors realize they can expand their market by peddling overpriced garbage to schools, cities, public transit systems, hospitals, etc -- which is how the "aggression detection" industry emerged, selling microphones whose "machine learning" backends are supposed to be able to detect "aggressive voices" (as well as gunshots) and alert cops or security guards. Read the rest

Laanlabs's showreel for 6d.ai meshing technology is an augmented reality demo in which virtual cockroaches crawl all over a very real kitchen. It's the best use of augmented reality I've ever seen. (via Beyond the Beyond) Read the rest

A new study from Facebook AI Research evaluates common machine-learning classifiers' ability to label photos of objects found in households in rich countries versus household objects from poor countries and finds that the models' performance lags significantly when being asked to classify the possessions of poor people. Read the rest

In Energy and Policy Considerations for Deep Learning in NLP, three UMass Amherst computer science researchers investigate the carbon budget of training machine learning models for natural language processing, and come back with the eyepopping headline figure of 78,468lbs to do a basic training-and-refinement operation. Read the rest

"The Training Commission" is Ingrid Burrington and Brendan C Byrne's serialized science fiction tale, taking the form of an email newsletter that lets you eavesdrop on the correspondence between the story's principal characters: it's set after a civil war ("the Shitstorm"), sparked by misbehaving and easily abused machine-learning systems, and which was resolved after a protracted and catastrophic digital blackout. Read the rest

Update: I've been emailed twice by Ever PR person Doug Aley, who incorrectly claimed that Ever's signup notice informed users that their data was going to be used to train an AI that would be marketed for military applications. It's true that during the signup process, users are asked whether they want to "use" facial recognition (that is, to label their images), but not whether they consent to having their images used to train that system, and especially not for commercial military applications.

Ever is an app that promises that you can "capture your memories" with unlimited photo storage, with sample albums featuring sentimental photos of grandparents and their grandkids; but Ever's parent company has another product, Ever AI, a facial recognition system pitched at military agencies to conduct population-scale surveillance. Though Ever's users' photos were used to train Ever AI, Ever AI's sales material omits this fact -- and the only way for Ever users to discover that their photos have become AI training data is to plough through a 2,500 "privacy policy." Read the rest

An adversarial preturbation is a small, human-imperceptible change to a piece of data that flummoxes an otherwise well-behaved machine learning classifier: for example, there's a really accurate ML model that guesses which full-sized image corresponds to a small thumbnail, but if you change just one pixel in the thumbnail, the classifier stops working almost entirely. Read the rest