Impression of the science done at OCCAM

Odessa Competence Center for Artificial intelligence and Machine Learning (OCCAM) is a private research laboratory founded by Dr. Arthur Franz and Michael Löffler in 2017 with the mission to advance both fundamental research and practical application of so-called artificial general intelligence (AGI). The name of the lab, OCCAM, is an allusion to the well-known reasoning principle, Occam's razor, stating that “entities should not be multiplied without necessity”. The laboratory is trying to make a contribution to humanity's long standing dream of constructing thinking machines that can solve a wide range of tasks without being specifically programmed for any of them.

Recently Arthur Franz, PhD, participated at the 12th conference on Artificial General Intelligence in Shenzhen, China. The AGI conference bestows prizes for exceptional contributions to the AGI field and this time Arthur Franz was awarded the Kurzweil Prize for the Best AGI Idea 2019 for the keynote talk:

“Toward an efficient AIXI approximation given the properties of our universe”.

The presented line of research contributes to the goal of formulating a practical theory of AGI, which in turn could allow researchers to actually implement it in future. The prize itself is named after Raymond Kurzweil, a famous American inventor, futurist and author, who pioneered several ideas in the fields of artificial intelligencs, robotics and transhumanism.

OCCAM’s goals are very ambitious but the founders see it rather as a challenge and not an obstacle:

“A key achievement of freedom is letting machines do the hard work. Machines however are usually not very intelligent. Artificial general intelligence (AGI) is the logical next step, in order to help us in areas where our minds are too limited. Only if we can make machines think and to be smarter than us, only then will we be able to ask them for help. Because we really need help in order to end suffering, finally and forever.”

Research projects:

Incremental Compression theory
OCCAM developed a theory of incremental compression, which can find short representations efficiently for arbitrary strings that can be generated by a composition of functions/features. Since data compression is crucial for AGI, this is a major step forward.

Theory of Hierarchial Compression
Incremental compression is an efficient way to find short representations for data generated by a composition of functions/features. Even though it is much more efficient than universal search, it is not enough for practical applications, since we don't know how to find those functions. However, since real world data usually show local correlations, this circumstance can be exploited by using a branching compression hierarchy.

Website | Facebook