Category Archives: Science

Learning Biofeedback Input System

By recording and analysing biofeedback and keyboard inputs, an analysis of the concurrent data could identify correlations to yield a biofeedback powered input system. The system would “learn” to replace the keyboard over time with better and better accuracy, faster input speed, and extended breadth of input types. Biofeedback devices could initially include medical biofeedback devices

Artificially Intelligent Program Trading System


Data Sources

/ / | | | \

Parallelized Analytics Agents

\ \ | | | / /

Signal Aggregation analytics

|

Webserver

|

Trading Platforms and Reporting

Arbitrages – even of very small marginal size – could be exploited based on a large number of artificially intelligent program trading systems mining historical and currently released information for financial products and correlated factors. Using multiple signalling agents, aggregated signals can be calculated by weighting each agent based on the statistical strength of historical signal combinations. Using this type of design, it is easy to develop multiple agents that are completely independent and confidential. Multiple agents can be developed concurrently, tested, and introduced to the aggregate system at any rate.

The process of the elimination of arbitrage opportunities will create vast concentrations of wealth within the companies that embrace the tools that automate this process. As new information sources become available for analysis, new arbitrages may be identified with increasing complexity. The abstraction of trading systems to automaically test and integrate new data sources will mark the last decades of financially advantageous investment in hedge funds. After that time, return will be a stochastic function of expected risk.

Self-Improving Image Recognition System

As image recognition is used and result sets are offered to users, the user selects the best fit result – and in doing so, indicates new data for future recognition. New angles, characteristics, and items could become recognizable with open training. Open training would consist of running the image recognition system on all available images (starting with web content, then uploadable images, then synchronization with networked cameras) and allowing any web user to click through the images by selecting the most appropriate result (or typing their own). After initial training, the system would be appropriate for many image recognition applications – with flexible adaptation to multiple specialized applications.

Entertainment is Evolving

Movies, music, images, and all sorts of digitally transferable media will migrate from broadcast to 2-way networks. Complex, and dynamically updating, personalization engines will determine content selection by default. The media will reach us over wired and wireless networks using highly compressed secure digital signals – rather than local storage. Centralized systems will have dominant personalization systems and data, and so will be the primary tool for selecting media and interface content. Media serving will be decentralized, however, as the media serving and the web serving are separated. Distribution will occur through all sorts of networked devices, including the personal computer… which will come to be characterized as a small wireless device used for all aspects of an individual’s computing needs. Mass media will evolve with the interface devices that are supported by personal computers. In other words, as our interface technology improves, mass media will be developed to take advantage of it. An example of this co-evolution will be the stereo monitor (one image for each eye) – allowing 3d graphics, and marking a key step for 3d media entertainment.