Ganbreeder uses a machine learning technique called Generative Adversarial Networks (GANs) to generate images that seem like photos, at least a first glance. Read the rest

CIT computer scientist Milan Cvitkovic conducted 46 in-depth interviews with "scientists, engineers, and CEOs" and collated their machine learning research needs into an aptly named paper entitled "Some Requests for Machine Learning Research from the East African Tech Scene," which presents an illuminating look into the gaps in the current practice of machine learning, itself an example of how rich-world priorities shape our ability to understand, compute and predict the world. Read the rest

Science fiction writer/lawyer Casey Fiesler is a maven in the field of tech ethics education (she maintains the amazing spreadsheet of tech-ethics syllabi); she uses science fiction stories as a jumping-off point for her own classroom discussions of ethics in technology. Read the rest

Researchers at NYU and U Michigan have published a paper explaining how they used a pair of machine-learning systems to develop a "universal fingerprint" that can fool the lowest-security fingerprint sensors 76% of the time (it is less effective against higher-security sensors). Read the rest

In 1975, Noam Chomsky and Jean Paiget held a historic debate about the nature of human cognition; Chomsky held that babies are born with a bunch of in-built rules and instincts that help them build up the knowledge that they need to navigate the world; Piaget argued that babies are effectively blank slates that acquire knowledge from experiencing the world (including the knowledge that there is a thing called "experience" and "the world"). Read the rest

Gwern Branwen asks the deceptively simple question "How many computers are in your computer?" Read the rest

Last year, Princeton researchers revealed a powerful new ad-blocking technique: perceptual ad-blocking uses a machine-learning model trained on images of pages with the ads identified to make predictions about which page elements are ads to block and which parts are not. Read the rest

InvaderZ is a Space Invaders variant that incorporates a genetic algorithm that mutates the invaders as you shoot at them, with survival for a fitness function: the longer an invader lasts before being blasted out of the sky, the more its behaviors are carried over into the next wave (here's a playable live version). (via Kottke) Read the rest

Frequent Boing Boing contributor Sean O'Brien and his colleagues Laurin Weissinger and Scott J Shapiro built a Raspberry Pi-enabled smart pumpkin and then challenged their Yale cybersecurity students to hack it. Read the rest

Deon is a project to create automated "ethics checklists" for data science projects; by default, running the code creates a comprehensive checklist covering data collection and storage, modeling and deployment: the checklist items aren't specific actions, they're "meant to provoke discussion among good-faith actors who take their ethical responsibilities seriously. Because of this, most of the items are framed as prompts to discuss or consider. Teams will want to document these discussions and decisions for posterity." Read the rest

Repairnator is a bot that identifies bugs in open source software integration and creates patches without human intervention, submitting them to the open source project's maintainers under an assumed human identity; it has succeeded in having five of its patches accepted so far. Read the rest

Robin Sloan is a programmer and novelist whose books like Sourdough and Mr Penumbra's 24-Hour Bookstore are rich and evocative blends of self-aware nerdy playfulness and magical speculation. Read the rest

Pete Warden (previously) writes persuasively that machine learning companies could make a ton of money by turning to data-compression: for example, ML systems could convert your speech to text, then back into speech using a high-fidelity facsimile of your voice at the other end, saving enormous amounts of bandwidth in between. Read the rest

Shitty math kills: shitty math "proved" that being selfish produced optimal outcomes and torched the planet; shitty math rains hellfire missiles down on innocents; in the 1960s, shitty math drove the "hamlet pacification program," producing 90,000 pages a month in data about the endless hail of bombs the US dropped on Vietnam. Read the rest

Joshua Browder launched DoNotPay when he started his computer science degree at Stanford; at first the app automated the process of fighting traffic tickets, then it expanded to helping homeless people claim benefits, then he automated suing Equifax for leaking all your financial data, then navigating the airlines' deliberately confusing process for getting refunds on plane tickets whose prices drop after you buy them. Read the rest

Yesterday, Bloomberg published a blockbuster story accusing the Chinese military of sneaking spy-chips "the size of a grain of rice" onto the motherboards of servers sold by Supermicro and/or Elemental for use in data-centers operated by the biggest US corporations (Apple and Amazon, among others), as well as US warships and military data-centers, and the servers used by Congress and the Senate. Read the rest

The heroic age of text adventure games was dominated by Zork and Zorkalikes, many from the games studio Infocom; the text adventures' fortunes sagged when improvements in computer graphics lowered the average gamer's age, and then rose again when BBSes carved new spaces for text-based play. Read the rest