Towards Telesophy: Federating All the World’ s Knowledge
Google engEDU
1 hr 6 min – Jul 11, 2007

Google Tech Talks
July 11, 2007

ABSTRACT

The Net is the global network, which enables users worldwide to interact with information. As new technologies mature, the functions of the protocols deepen, moving closer to cyberspace visions of "being one with all the world’s knowledge". The Evolution of the Net has already proceeded from data transmission in the Internet to information retrieval in the Web. The global protocols are evolving towards knowledge navigation in the Interspace, moving from syntax to semantics. In the future, infrastructure will support analysis, for interactive correlations across knowledge sources. This moves closer towards "telesophy", (transparent infrastructure for) knowledge at a distance.

Getting from here to there will require a paradigm shift from central to distributed, from searching universal archives to navigating community repositories. Central archives partially survived the transition from a million repositories to a billion, but distributed indexing is necessary to scale to a trillion repositories in the next generation. Supporting scalable semantics requires divide-and-conquer to capture local context as an approximation to global meaning. Concept switches in the Interspace are the analogue of packet switches in the Internet, since user interaction is at the level of logical spaces rather than physical networks. This talk will describe the research technologies and trends creating the global infrastructure, with suggestions for hero experiments and hints at the new world of the near future.

Speaker: Bruce Schatz
Bruce Schatz is Director of the CANIS (Community Architectures for Network Information Systems) Laboratory at the University of Illinois at Urbana-Champaign. He has been working on federating all the world’s knowledge by building pioneering research systems in industrial and academic settings. These include the first large-scale prototype for federating syntax (Telesophy at Bellcore in the 1980s), structure (DeLIver at UIUC in the 1990s), and semantics (BeeSpace at UIUC in the 2000s). During the 1980s, he was the member of the Internet Research Task Force focused on search and during the 1990s was the Principal Investigator of the Illinois project in the NSF Digital Libraries Initiative, the search flagship in the federal program that spun off Google. Read the rest of this entry »

2,3,5, Infinity!
Google engEDU
52 min – Aug 18, 2006

Google Tech Talks
August 18, 2006

Paul Hildebrandt founded Zometool Inc. with co-inventor Marc Pelletier in 1985. He organized the Zome team, raised capital and coordinated research and development of prototype tooling, production systems, packaging and collateral material, and continues as president and board chairman. He graduated magna cum laude in economics from the University of Colorado (1987); was Director of Research for Eco-Cycle, Inc., largest local recycling program in the U.S. (1982–1985); and founded Renaissance Community Press, a print shop for human service organizations (1976–1982). Several times he’s crossed the US by freight train for less than $20, and has ceased having out-of-body experiences since getting married in 1992 or ‘93.

ABSTRACT
Nearly 60 years after the first electronic digital computer was designed at the Princeton Institute for Advanced Studies (IAS), companies like Google are demonstrating the power of a world built from 1s and 0s. Zome is a system that models the world built from the numbers 2, 3 and 5. We will explore how these numbers are knotted together to form the structure of space, from the subatomic framework of the atom, to the geometry of life, to a recently proposed “shape” of the universe! Read the rest of this entry »

The Implications of OpenID
Google engEDU
51 min – Jun 25, 2007

Google Tech Talks
June 25, 2007

ABSTRACT

Simon Willison
OpenID is an emerging standard that provides simple, decentralised authentication for the Web. OpenID follows the Unix philosophy, solving one small problem rather than attempting to tackle the many larger challenges posed by online identity. This talk will explore the implications of OpenID, and explore the best practices required to take advantage of this new technology while avoiding the potential pitfalls.

Speaker: Simon Willison
Simon Willison is a consultant on OpenID and client- and server-side Web development, and a co-creator of the Django Web framework. Before going frelance Simon worked on Yahoo!’s Technology Development team, and prior to that at the Lawrence Journal-World, an award winning local newspaper in Kansas. Simon maintains a popular Web development weblog at http://simonwillison.net/ Read the rest of this entry »

Scalable Learning and Inference in Hierarchical Models of the Neocortex
Google engEDU
53 min – Jan 17, 2006

Google TechTalks
January 17, 2006

Tom Dean

ABSTRACT
Borrowing insights from computational neuroscience, we present a class of generative models well suited to modeling perceptual processes and an algorithm for learning their parameters that promises to scale to learning very large models. The models are hierarchical, composed of multiple levels, and allow input only at the lowest level, the base of the hierarchy. Connections within a level are generally local and may or may not be directed. Connections between levels are directed and generally do not span multiple levels.

The learning algorithm falls within the general family of expectation maximization algorithms. Parameter estimation proceeds level-by-level starting with components in the lowest level and moving up the hierarchy.

The inference required for learning is carried out by local message passing and the arrangement of connections within the underlying networks is designed to facilitate this method of inference. Learning is unsupervised but can be easily adapted to accommodate labeled data. Read the rest of this entry »

Automated Reconstruction of 3D Models
Google engEDU
1 hr 2 min – Oct 25, 2006

Google Tech Talks
October 25, 2006

Avideh Zakhor, Prof. UC Berkeley

ABSTRACT
In this talk, we present a fast approach to automated generation of textured 3D city models with both high details at ground level, and complete coverage for birds-eye view. The goal is photorealistic rendering for walk throughs, drive through and fly throughs. A close-range facade model is acquired at the ground level by driving a vehicle equipped with laser scanners and a digital camera under normal traffic conditions on public roads in a continuous, rather than a stop-and-go fashion, resulting in extremeley fast data acquisition times; a far-range Digital Surface Model (DSM), containing complementary roof and terrain shape, is created from airborne laser scans, then triangulated, and finally texture-mapped with aerial imagery. The facade models are first registered with respect to the DSM using Monte-Carlo-Localization, and then merged with the DSM by removing redundant parts and filling gaps. The continuous mode scanning, combined with a no human in the loop approach, has enabled us to generated detailed models of downtown Berkeley facades with 25 minutes of driving under normal traffic conditions and 4 hours of automated processing on a single CPU personal computer. We will show the resulting downtown Berkeley models using both commercial vrml viewers, as well as inserted in Google Earth. Read the rest of this entry »