the problems of NLP are highly interconnected are need to be solved as a whole, yet very few efforts have been made to attack several frontiers at once. most research is very academic and niche-oriented. assertion is that CPU power and amounts of data, such as wikipedia, will create breakthroughs in NLP.
peter starts with the well-known globe with the google queries superimposed. “google saves over 9000 man-years of effort every year. So Google saves 9000 lives per year.” numerous mentions of people making a living from google ads. shows keyhole. “the computers from star trek are always omniscient but never helpful, they never tell you: don’t do this.” the spelling checker is not dictionary based but works on their huge accumulated data, like the 500 spellings of britney. the web is a million times larger than the largest computational linguistics corpuses. a billion documents really makes the difference according to banko and brill. humans achieve around 95% accuracy. work is being done on semantic understanding. one program they run is extracting categories and members of these categories from their corpus. “done very simply: you take the whole web, break it into sentences, and look for about half a dozen patterns, such as including, as in Software companies, including..” they take an automated approach to machine translation by looking for pages on the web that have documents in two languages and derive the model from it. this yields a level of translation that is good enough. “doug cutting was more interested in the crawling and indexing side, so thats where lucene looks good, not so much in the sorting.” google focuses on the 95%, the easy part, to get more leverage, but will have to go back to the hard part. they found that the feedback button didn’t work. some people were writing them that they were looking for a specific book by typing in “library”, which indicates a deeper problem. the first google logo was done during burning man 1999 to indicate that “no one is at the office, don’t blame us”.
peter thiel was one of the VC’s behind paypal. he starts of with “why has money value?” and expands it to “what are the things that have value in an economy?”. more and more of the value of an economy are in intangibles.. asserts that john locke started the march to VR by describing nature as cruel and the desire to create an alternate world. one of the reasons for setting up paypal was to create payment systems in cyberspace that were as independent from the real world. “50% of world wealth is now stored in offshore locations.” peter mentions sealand, the “virtual country”. i don’t exactly got his point to be honest.
richard marks just showed a couple motion tracking demos they developed for eye toys. google has images. some of the demos were so awesome that it drew loud cheering from this jaded crowd. farther out are sensors with depth detection which allows to have truly 3D integration, where you can “reach out” to interact with the system. the demo shows butterflies circling around a guy as he walks in front of the camera. this is truly awesome. also, for the DDR crowd, it now supports eye toys.
doug engelbart requires no introduction. his 1968 demo is legendary, and here is to hoping he can teach us young upstarts some tricks to bootstrap. doug moved to the bay area in 1948 and had a life-changing epiphany that he had no goals in his career. so he settled on a goal to maximize his contribution to mankind. his goal: to boost mankinds collective ability for coping with complex, urgent problems. at the time, there were about 3 computers in the US. doug starts talking about CODIAK capabilities, for Concurrent Development, Integration, and Application of Knowledge. he asserts that human cognitive abilities depend on their augmentation system. for each tool system, a human system is changed: Media vs Perception, Portrayal vs Organization etc. Their original system had a steep learning curve because it focussed on concepts such as transposing and groups while traditional office automation called for howto-like, step by step instructions. scaling of augmentation is a whole science to itself: Dimensionality Numbers. The example doug used was that if you were a thousand times heavier you would not be a thousand times stronger, but only a hundred times because strength scales differently. Which would be the same as having a tenth of your strength with your current weight
Doug also mentions a sample of the increasing complexity of world problems: The State of the Future which appears every year. (2004 report)
A definition of boostrapping: “Pursue the capabilities that increase the capabilities to pursue.”
most interesting question: “online games creation inflation (with monsters being money trees basically). how can you keep that inflationary pressure out if you trade game items with real cash?”
will wright, the creator of the sims, is speaking now. will is showing funny episodes from his online world “here i am hanging out with my boyz, enjoying refreshments” (pictures of a bad neighborhood on the screen). now he talks about the concept of possibility space. (reminds me of the all possible universes discussion). “you can describe the meaningful verbs of a game very easily: kill, race, manage.” “we found that user were quickly comfortable with the idea of shopping, it soothed them.” will asserts that science has been moving away from calculus to simulation, while games have gone in the opposite direction. thus the creation of elaborate physics engines. now he is showing lots of interesting slides in quick succession about feedback mechanisms in games, topologies, and patterns. he is talking far too quickly for me to summarize it. “failures are entertaining, it becomes boring quickly to just explore the path of success”. Cybernetics, System Dynamics, Cellular Automata and Chaos Theory are some approaches to model reality that have been developed. Will also mentions the old favorite of the blogosphere, the Power law distribution. “Relativity theory and quantum mechanics describe different aspects of the universe (very successfully), but taken together they cannot explain a duck.” Game development is moving from 1 person efforts in 1984 to several hundreds for The Sims 2. Extrapolating from that quickly gets you $500M games with 2.5M developers. “People are increasingly designing their environments: Cell phone covers, fashion, interiors..” Machinima is becoming increasingly popular: shooting film in virtual reality. In the future, metrics are becoming important for the further development of these games. Looking at purchase history, relationships allow to devise elaborate model of user behavior. Based on these patterns, potential matches can be found (think tickle on steroids). someone asks about copyright protection, and will says he personally uses cracks “its the easiest.”
You really need to see his presentation on DVD to get the full effect, i am trying to snag a copy.
my new tshirt source.
another tidbit: someone mentioned that countries like india value engineer types as highly as other countries value the jock. the nerds get all the girls according to this correspondent because parents tell their kids thats where the money is. it would be interesting to see this substantiated with a study. of course, paul graham is on it.
dewayne hendricks is talking about CENIC’s One Gigabit or Bust™ Initiative which addresses critical technical, policy, financial and organization challenges facing the delivery of one gigabit broadband to all Californians by 2010. i am kinda fried from all the cool talks, so just a sound bite: “by 2010, we will have 1GB wireless equipment just as we have wifi today”