|
|
about us - FAQ |
A Paradigm Shift in the Development of Intelligent Systems - Artificial Innovation
Industry currently makes extensive use of a variety of complex software systems, supported by increasingly powerful computational hardware, to design and develop cars, planes, spacecraft, and even drugs. But the constraints of our knowledge are built into the software design.
There is an underlying assumption that we know how to solve the problems we are faced with (i.e., we can develop mathematical algorithms to deliver the "right answers" to questions); we just need the power of today's supercomputers to speed the execution of the complex calculations. However, humans do not actually have the knowledge to solve some of the most complex problems that face us, so how can we hope to write programs to provide us with the solutions? For example, humans can think in 3-dimensions, but we know that space has a fourth dimension - time - and yet we can not work with it.
Humans tend to approach problem-solving in a sequential manner, breaking down the problem into its components (or dimensions) and analyzing / solving each one before moving onto the next. This is the approach typically used in drug discovery.
Matrix Pharma has developed its own computational strategies and theories that let the software system do the innovative and creative parts. The software system "evolves", automatically inventing the laws and theories required to solve complex problems without any human intervention or programming.
Below, we try to answer a few of the questions that are asked most often:
How does the software work?
Actually, it is not one piece of software; it is a system of software components adapted to the special needs of each problem. The basis of the system is a module that starts with random computer code (yes, this is possible!). The random code is modelled around the data like modelling clay around some given form. The code is executed and - as is standard scientific practice - challenged with the real-world problem for further optimization and verification
What is the basis for the software system?
The modelling procedure works within the syntax and semantics of the programming language that is used. Nothing more is needed. So the software programs itself without any constrains of pre-existing formulas or theories. It is unaware of the constraints that may or may not exist.
Does the process truly start from nothing, that is, without any assumptions or hypotheses?
Yes
Can it really develop theories that go beyond human knowledge, and which can be proven scientifically?
Yes, and this has been accomplished many times in collaboration with universities, companies, and other institutions.
What can this software already do today and has proven in the past that it can do it?
The company has delivered over 100 projects over the last years.
Examples:
Drug Development: Discovered drugs for thrombosis (licensed to BiolineRx), antibiotics, and even fungicides on whole living plants.
Patient Data Analysis: Analysed patient data such that new treatments or diagnostics can be derived from normal blood analysis parameters.
Diagnostics: Extracted from biomarker data, small non linear networks that are capable of identifying diseases much more precisely than other methods currently in use.
Quality Control: Analysed the entire production process of a company using instrument and factory sensor data to assess quality level of product at various production stages. Errors in quality could be predicted.
Invention of New Picture Compression Algorithm: Developed a completely new algorithm to compress specific types of pictures of a particular quality with a given (predefined) compression factor.
For more information about Artificial Innovation or if you have any more questions please contact us.
|
|
