- Find The Future — New York Public Library big game, by Jane McGonigal. (via Imran Ali)
- Enable Certificate Checking on Mac OS X — how to get your browser to catch attempts to trick you with revoked certificates (more of a worry since security problems at certificate authorities came to light). (via Peter Biddle)
- Clever Algorithms — Nature-Inspired Programming Recipes from AI, examples in Ruby. I hadn’t realized there were Artificial Immune Systems. Cool! (via Avi Bryant)
- Rethinking Evaluation Metrics in Light of Flickr Commons — conference paper from Museums and the Web. As you move from “we are publishing, you are reading” to a read-write web, you must change your metrics. Rather than import comments and tags directly into the Library’s catalog, we verify the information then use it to expand the records of photos that had little description when acquired by the Library. [...] The symbolic 2,500th record, a photo from the Bain collection of Captain Charles Polack, is illustrative of the updates taking place based on community input. The new information in the record of this photo now includes his full name, death date, employer, and the occasion for taking the photo, the 100th Atlantic crossing as ocean liner captain. An additional note added to the record points the Library visitor to the Flickr conversation and more of the story with references to gold shipments during WWI. Qualitative measurements, like level of engagement, are a challenge to gauge and convey. While resources expended are sometimes viewed as a cost, in this case they indicate benefit. If you don’t measure the right thing, you’ll view success as a failure. (via Seb Chan)
ENTRIES TAGGED "ai"
Four short links: 4 April 2011
Library Game, Mac Security, Natural Programming, Selecting Metrics
Four short links: 21 February 2011
Data in Javascript, Artificial Empathy, Gender Comms, and Artificial Scarcity
- Amplify.js — simplify all forms of data handling by providing a unified [Javascript] API for various data sources. Amplify’s store component handles persistent client-side storage, using standards like localStorage and sessionStorage, but falling back on non-standard implementations for older browsers. Amplify’s request adds some additional features to jQuery’s ajax method while abstracting away the underlying data source.
- Artificial Empathy (Matt Jones) — we’ve evolved to broadcast and receive emotion, to infer intelligence and intent from the weakest of signals. Now we’re starting to use those communication channels in computer interactions. Matt Jones wonders what we can learn from animals about how this will play out.
- Communication Styles Make a Difference (NY Times) — Women were also more negative about the tone of the list. Whereas men tended to say that they found the “slings and arrows” that list members posted “entertaining” (as long as they weren’t directed at them), women reported that the antagonistic exchanges made them want to unsubscribe from the list. One women said it made her want to drop out of the field [...] altogether. Also interesting: neutral point-of-view penalizes anyone who cautiously phrases contributions as opinion and rewards those who boldly claim facthood.
- Why Platforms Leak: The Impact of Artificial Scarcity (JP Rangaswami) — the best summary of the inevitable failure of artificial scarcity. When you make something digital and connect it to the web, it becomes available everywhere, it becomes available immediately [...] As we’ve moved from the physical world to the digital world, incumbents in many industries have sought to preserve the historical structures and ways of doing business. Which, in effect, were attempts to create and exploit artificial scarcities. When it comes to digital assets, there are four primary ways to try and create artificial scarcity. [...] All these have been attempted. All these have failed, and will continue to fail. You cannot make something that is essentially abundant artificially scarce.
Four short links: 3 Sep 2010
Design Principles, Mario AI, Open Source Wave, and 3D Google Earth Sound
- Arranging Things: The Rhetoric of Object Placement (Amazon) — [...] the underlying principles that govern how Western designers arrange things in three-dimensional compositions. Inspired by Greek and Roman notions of rhetoric [...] Koren elucidates the elements of arranging rhetoric that all designers instinctively use in everything from floral compositions to interior decorating. (via Elaine Wherry)
- 2010 Mario AI Championship — three tracks: Gameplay, Learning, and Level Generation. Found via Ben Weber’s account of his Level Generation entry. My submission utilizes a multi-pass approach to level generation in which the system iterates through the level several times, placing different types of objects during each pass. During each pass through the level, a subset of each object type has a specific probability of being added to the level. The result is a computationally efficient approach to generating a large space of randomized levels.
- Wave in a Box — Google to flesh out existing open source Wave client and server into full “Wave in a Box” app status.
- 3D Sound in Google Earth (YouTube) — wow. (via Planet In Action)
Watson, Turing, and extreme machine learning
The real value of the Watson supercomputer will come from what it inspires.
While IBM's Watson supercomputer / Jeopardy contestant is a masterpiece of natural language processing, it's important to remember that it's just a learning tool that will help us solve more interesting problems.
Four short links: 17 June 2010
Statistical Jeopardy Wins, Mobile Taxonomy, Geodata Mystery, and Machine Learning Blog
- What is IBM’s Watson? (NY Times) — IBM joining the big data machine learning race, and hatching a Blue Gene system that can answer Jeopardy questions. Does good, not great, and is getting better.
- Google Lays Out its Mobile Strategy (InformationWeek) — notable to me for Rechis said that Google breaks down mobile users into three behavior groups: A. “Repetitive now” B. “Bored now” C. “Urgent now”, a useful way to look at it. (via Tim)
- BP GIS and the Mysteriously Vanishing Letter — intrigue in the geodata world. This post makes it sound as though cleanup data is going into a box behind BP’s firewall, and the folks who said “um, the government should be the depot, because it needs to know it has a guaranteed-untampered and guaranteed-able-to-access copy of this data” were fired. For more info, including on the data that is available, see the geowanking thread.
- Streamhacker — a blog talking about text mining and other good things, with nltk code you can run. (via heraldxchaos on Delicious)
Four short links: 25 March 2010
Against Open Data, Singalong Selection, Library Release, and Twitter Analysis
- Aren’t You Being a Little Hasty in Making This Data Free? — very nice deconstruction of a letter sent by ESRI and competitors to the British Government, alarmed at the announcement that various small- and mid-sized datasets would no longer be charged for. In short, companies that make money reselling datasets hate the idea of free datasets. The arguments against charging are that the cost of gating access exceeds revenue and that open access maximises economic gain. (via glynmoody on Twitter)
- User Assisted Audio Selection — amazing movie that lets you sing or hum along with a piece of music to pull them out of the background music. The researcher, Paris Smaragdis has a done lot of other nifty audio work. (via waxpancake on Twitter)
- Cologne-based Libraries Release 5.4M Bibliographic Records to CC0 — I see resonance here with the Cologne Archives disaster last year, where the building collapsed and 18km of shelves covering over 2000 years of municipal history were lost. When you have digital heritage, embrace the ease of copying and spread those bits as far and wide as you can. Hoarding bits comes with a risk of a digital Cologne disaster, where one calamity deletes your collection. (via glynmoody on Twitter)
- ThinkTank — web app that lets you analyse your tweets, break down responses to queries, and archive your Twitter experience. Built by Expert Labs.
Four short links: 3 March 2010
Most Dangerous Code, 3D Printing, Artificial Irony, and Live Visualisation
- Top 25 Most Dangerous Programming Errors (MITRE) — I could play bingo with this on some of the programs I wrote when I was learning to code. Now, of course, I am perfect. *cough*cough*
- RepRap Printing in Clay — interesting because of the high price of the plastic that fab units typically use. Other groups are working on this–see, for example, recycled glass, sugar, and maltodextrin.
- Artificial Flight and Other Myths — amusing parody of anti-AI arguments.
- Snake Oil Supplements — visualisation of the scientific evidence for various food supplements. What interested me is that it’s automatically generated from data in this Google Doc.
Four short links: 24 December 2009
Minds for Sale, Heat Death of the Web, Handheld Wireless Magazines, Joking Computers
- Jonathan Zittrain on “Minds for Sale” — video of a presentation he gave at the Computer History Museum about crowdsourcing. In the words of one attendee, Zittrain focuses on the potential alienation and opportunities for abuse that can arise with the growth of distributed online production. He also contemplates the thin line that separates exploitation from volunteering in the context of online communities and collaboration. Video embedded below.
- Anatomy of a Bad Search Result — Physicists tell us that the 2nd law of Thermodynamics predicts that eventually everything in the universe will be the same temperature, the way a hot bath in a cold room ends up being a lukewarm bath in a lukewarm room. The web is entering its own heat death as SEO scum build fake sites with stolen content from elsewhere on the web. If this continues, we won’t be able to find good content for all the bullshit. The key is to have enough dishwaster-related text to look like it’s a blog about dishwashers, while also having enough text diversity to avoid being detected by Google as duplicative or automatically generated content. So who created this fake blog? It could have been Consumersearch, or a “black hat” SEO consultant, or someone in an affiliate program that Consumersearch doesn’t even know. I’m not trying to imply that Consumersearch did anything wrong. The problem is systematic. When you have a multibillion dollar economy built around keywords and links, the ultimate “products” optimize for just that: keywords and links. The incentive to create quality content diminishes.
- Magplus — gorgeous prototyping for how magazines might work on new handheld devices.
- Glasgow’s Joking Computer — The Glasgow Science Centre in Scotland is exhibiting a computer that makes up jokes using its database of simple language rules and a large vocabulary. It’s doing better than most 8 year old children. In fact, if we were perfectly honest, most adults can’t pun to save themselves. Q: What do you call a shout with a window? A: A computer scream. (via Physorg News)
Radar 
Radar on 
Radar on 
Radar on 
Radar on 