Nat has chaired the O'Reilly Open Source Convention and other O'Reilly conferences for over a decade. He ran the first web server in New Zealand, co-wrote the best-selling Perl Cookbook, and was one of the founding Radar bloggers. He lives in New Zealand and consults in the Asia-Pacific region.
Giphy Wants All the GIFs (Backchannel) — they’re licensing legit high-res metadata-rich GIFs from content producers, wanting to become a real search engine. The description of metadata problems on user contributions is interesting. GIF tags were incredibly spammy (“every teenage kid tags everything ‘One Direction’ and ‘Bieber,’” gripes Chung), and many of the GIFs themselves were low-quality due to file-size limits on popular sites such as Reddit and Tumblr. And there was a ton of porn. They were forced to manually clean it up, adding good metadata and using Mechanical Turk to get rid of the NSFW files.
Mika Model (Slate) — a new story from Paolo Bacigalupi, asking whether an AI’d up sex doll could commit murder. The companion article by Ryan Calo takes the story’s situation seriously and tries to answer that question, and is fascinating.
Strategic Implications of Openness in AI Development (Nick Bostrom, PDF) — Some forms of openness are plausibly positive on both counts (openness about safety measures, openness about goals). Others (openness about source code, science, and possibly capability) could e.g. lead to a tightening of the competitive situation around the time of the introduction of advanced AI, increasing the probability that winning the AI race is incompatible with using any safety method that incurs a delay or limits performance. We identify several key factors that must be taken into account by any well-founded opinion on the matter. An argument for a closed-source singularity.
LimeSDR — a low cost, open source, apps-enabled (more on that later) software defined radio (SDR) platform that can be used to support just about any type of wireless communication standard. With an App Store (via Snappy Ubuntu Core’s app distribution platform).
NHS Data Sharing (Ben Goldacre) — as this project lands, we’re all becoming rapidly aware that incompetence, malice and creepiness around confidential data is policed with a worryingly light touch. Ben is scrupulously fair with the benefits and the risks, as clear on the state-of-the-art limitations for data sharing as the opportunities. We have a golden opportunity in the UK, with 60 million people cared for in one glorious NHS.
Underactuated Robotics — MIT coursenotes, forming a working draft of a book is about building robots that move with speed, efficiency, and grace. I believe that this can only be achieve[d] through a tight coupling between mechanical design, passive dynamics, and nonlinear control synthesis. Therefore, these notes contain selected material from dynamical systems theory, as well as linear and nonlinear control.
April Sees Robotics Investment Up (Robohub) — April was a big month for investing in robotics – 19 companies were funded to the tune of $175 million vs. $15.8M in January, $18.6M in February, and $45.4M in March. Four companies were acquired with 3 of the 4 reporting selling prices totaling $422 million.
Inevitability in Technology (Ben Evans) — We think of the portal model as a dead-end, but half a billion Chinese Internet users suggest that it could have been otherwise. The Chinese internet is a great way to challenge your thinking on what’s inevitable in technology – it’s a living counterfactual.
Digital Genies — The worst thing is a machine that has the wrong values, but is absolutely convinced it has the right ones, because then there’s nothing you can do to divert it from the path it thinks it’s supposed to be following. But if it’s uncertain about what it’s supposed to be following, a lot of the issues become easier to deal with because then the machine says, OK, I know that I’m supposed to be optimizing human values, but I don’t know what they are. It’s precisely this uncertainty that makes the machine safer, because it’s not single minded in pursuing its objectives. It allows itself to be corrected.
Twitter’s AI for Live Video (MIT TR) — deep learning to recognise activity and objects in videos, so Periscope live streams can be searched for and found. I imagine there’s a large portion of the deep learning devoted to forbidden pink.
Well-Crafted Websites (Adrian Holovaty) — It’s clear to me, studying the history, that service workers are the next big “well-crafted” hint. The signs and similarities are all there. Today, they’re used mostly by fringe sites, in an experimental fashion. It’s inevitable that in a short generation they’ll become common and expected.
OpenAI Gym — A toolkit for developing and comparing reinforcement learning algorithms. It supports teaching agents everything from walking to playing games like Pong or Go.
Open Xamarin — the Xamarin tools are now all MIT-licensed, after the Microsoft acquisition. And show no signs of being abandoned.
Text-Mining the History of Medicine — In this article, we present our efforts to overcome the various challenges faced in the semantic analysis of published historical medical text dating back to the mid 19th century.
A Journal Is a Club (Cameron Neylon) — I’ve been frustrated for a long time with traditional economic analyses of scholarly publishing. They don’t seem to explain what actually happens, and fail to capture critical aspects of what is going on. The lens of club economics seems like it might help to capture more of the reality of what is going on.
The Shape of Things (Tom Coates) — In fact it’s this problem of what’s most intuitive that gives me most pause for tangible computing generally. The assumption from many of these thinkers is that making an interface that’s physical makes it inherently more intuitive. But I don’t buy that physical affordances alone will make it immediately obvious what a smart connected object is for. Sure, you pick up a hammer and you immediately want to hit something (or maybe that’s just me) — but is that true of a smart hammer?
Improving the Peer Review Process: A Proposed Market System (PDF) — We thus suggest a more efficient and integrity-preserving system based on an open two-sided market in which buyers and sellers of peer review services would both be subject to a set of recursive quality indicators. We lay out key features we think would be important to reduce the opportunities for gaming and that improve the signals about the societal value of a contribution. Cool story bro, but until academics are rewarded financially/professionally for publishing in a Better System, most will accept/route-around the current system.
Metaphors of Data, a Reading List — The goal in assembling this list was to catalog resources that are helpful in unpacking and critiquing different metaphors, ranging from the hype around data as the new oil to less common (and perhaps more curious) formulations, such as data as sweat or toxic waste.
Routersploit — metasploit-like framework for banging on routers.
Uncanny Valley — “This is the next big company,” he had said. “It’s a rocket ship.” He was right. I had been banking on him being right. Still, there are days when all I want is to disembark, eject myself into space, admit defeat.
Driverless Trucking Numbers (TechCrunch) — $4.5k to truck something across country, 75% of which is labour. Trucks most fuel-efficient at 45mph but drivers are paid by the hour and their hours capped at 11/day. More truck drivers killed on the job than any other occupation. Truck drivers are 1% of the workforce, and it’s the most common job in 29 states. And more “gosh this is gonna be interesting” numbers.
Email Isn’t the Problem — Glyph nails it. The thing you are bad at is saying ‘no’ to people. The downside of lauding 20 year olds in tech is that they build software for other 20 year olds: software that creates noise, distraction, opportunity. When you are doing what you want to do, though, that same software inhibits your ability to do it. All we have is poorly-evolved meat to fight the wily silicon ….
Terrible Truths of Pricing — The first rule of pricing is that you don’t talk about pricing. What he says might be true, but the absence of any question of morality around pricing drugs (which he tackles outright) makes me grumble “price of everything and value of nothing”.
Boaty McBoatface and the False Promise of Democracy — my take: you want opinions, but you also want committed opinions. Your poll/survey/vote will erect (or fail to erect) barriers to participation, and those barriers represent a measure of commitment. No barriers = lots of votes, but high risk of Boaty McBoatface. High barriers = few votes, but from those who care.
Team Models for Scaling a Design System — actually a dig into federated design teams, which is the rough shape that a lot of large companies end up in. The fundamental tension: As a design system stabilizes, every team I’ve observed has designers that do build, document and sustain it, and designers that can’t or say they want to but don’t.
Anthropic Capitalism And The New Gimmick Economy — market capitalism struggles with “public goods” (those which are inexhaustible and non-excludable, like infinitely copyable bits that any number of people can have copies of at once), yet much of the world is being recast as an activity where software manipulates information, thus becoming a public good. Capitalism and Communism, which briefly resembled victor and vanquished, increasingly look more like Thelma and Louise; a tragic couple sent over the edge by forces beyond their control. What comes next is anyone’s guess and the world hangs in the balance.