The Technology of CorticalDB April 27, 2009 ( by Thom Lynch)
Even the brain of a small mammal like a mouse or rat exceeds the possibilities of the largest computers in the world.
Scientists have now been working for years in an attempt to determine exactly how the brain could be emulated and used in computing or artificial intelligence.
It seems the stuff of science fiction. Yet today, neural networks based on a basic understanding of how mammalian brains process information is common place. More recently, Spiking Neural Networks have been used to bring even more to the field. Powerful new algorithms based on this technology are now commonplace. But a real emulation of a brain has not yet been successfully achieved.
We first started working on this very kind of challenge back in 2003. How exactly could we achieve mammalian brain simulation without having to use the largest computer in the universe? In fact, IBM was trying to do just that very thing with massive amounts of processing power and CPU space. But that way is slow, expensive, and seemingly never ending.
We put together a research team in India to study and capture the data needed to get to where we needed to be. We hired programmers, mathematicians, and computer infrastructure professionals and set them to work, while also developing our own algorithms to design and efficiently run large scale cortical simulations.
These algorithms have been coded in C, using all-in-memory associative arrays, search-engine style inverted indexes, and dictionary techniques similar to those used in the Lempel-Ziv-Welch compression algorithm. The result is CCortex, a powerful set of algorithms and software libraries intended for very fast cortical emulation, algorithms packaged as the CCortex software framework intended to design and efficiently run large scale cortical simulations.
Our first thoughts were to take this exciting knowledge base that we have been building up for more than 5 years and license CCortex™ to other developers. The opportunities are nearly endless; with application to security, marketing, retail, database, and other any other industry that could benefit from improved understanding of unstructured data. However, while working on the research, the company itself began to look at how it could develop its own software applications.
This blog will be a way that we can keep you informed on our progress and invite you to share your thoughts about our products, the field in general, or nearly anything else you might want to discuss.
Our Heart is in San Francisco February 14, 2009 ( by Thom Lynch)
We moved our headquarters from Palo Alto to San Francisco to take advantage of San Francisco's world class communications and infrastructure as we become focused on providing enterprise-level support and services. This was a natural outgrowth as it became clearer to us that we needed to shift our focus from being a purely Research and Development enterprise, to a technology solutions provider. When the company was solely R&D-focused, our proximity to the rich academic community in Palo Alto was crucial to our development.
The new headquarters will house our leadership team, company administration, and sales and marketing departments, while much of the research and development teams will remain in India and Europe.
The company is changing its name to CorticalDB to reflect the shift in focus and to follow up on decisions about which products to focus on first using our exciting core technologies.
We'll have a lot more to share with you in the future about our own internally developed products. We still plan to license our technology to other developers as we always have, but think that we also have some great ideas for products that we can develop.
CorticalDB Releases More Information about Products in Development
We've set up a new website to keep you up to date at what is happening at CorticalDB. We've announced, with the release of the website, the two core products that we'd like you to know about. You'll find those areas each have their own section on the site.
CorticalDB will provide the technology and systems for providing and servicing all of our other products and those of our development partners. These applications include:
Advanced Enterprise Search: Cortical Systems not only uncovers, but also makes sense of, the 85% of enterprise information that is hidden to all other technologies including keyword search engines and relational databases. As a result, users are presented with relevant information they didn't even know existed so they can act on it in real-time.
Knowledge Management: Cortical Systems enables organizations to automatically form a contextual understanding of people's interests, behavior and ongoing interaction with any type of information. This facilitates collaboration by leveraging the most valuable knowledge available the experience and expertise of an organization's employees.
Electronic Discovery: Cortical Systems enables organizations to extract meaningful evidence from terabytes of email, documents, spreadsheets and other unstructured information. This enhances the ability of investigators to make informed decisions about corporate culture and track the development of illegal activity.
CCortex Core Libraries are a set of tools and software libraries, optimized to enable the deployment of sophisticate neural networks. CCortex effectively mimics the structure of the human cortex, the outer layer of gray matter at the cerebral hemispheres, largely responsible for higher brain functions. CCortex technology is the centerpiece of all CorticalDB products.
Additionally, you'll find a section called the CorticalDB LAB where you can keep up on the work, thoughts, improvements, and more from us and our development partners. This is a great way to see the power of the CorticalDB and the broad number of applications that can make it work.