The startup SkyOptim stems from a friendship between two individuals who, through their discussions,
discovered a new approach to thinking about Artificial Intelligence.
The fundamental idea is that Artificial Intelligence is innovative, and it is this aspect that
sets it apart from statistics. To clarify this point, one only needs to observe how chess programs based
on Deep Learning have succeeded in innovating. They have developed strategies and playing styles that
were previously unknown. What is fascinating to observe is that these programs were trained on millions
of games played in the past and managed to surpass the strategies contained in these games, meaning they
devised different ways of approaching the game.
This ability to innovate fundamentally distinguishes Artificial Intelligence from statistics. A tool
based on statistics, trained with a knowledge base, merely reproduces patterns found in that
base and lacks the capacity to innovate.
It is interesting to note that this characterization of Artificial Intelligence is actually very ancient
in human history. Referring to the art of combat, various Asian martial arts, such as Tai Chi Chuan,
follow the same principle. The idea is to open one's body to be able to move it in all possible
directions, producing an "innovative" gesture, one that has not been made by the body in the past. In
doing so, the martial artist becomes "undetectable," meaning one cannot anticipate their actions by
examining their posture. In the art of the Japanese sword, the same idea is found, with specific
training designed to free the body from its conditioning and promote the "innovative gesture," one that
arises without being slowed down by the mechanism of thought.
Our two friends shared this view of martial arts and dedicated their leisure time to designing
algorithms that exploited this aspect. It is from this shared passion that SkyOptim was born, a startup
focused on analyzing air traffic operations to identify "innovation opportunities" and associated gains.
Founder of SkyOptim
Sasha has a remarkable career journey, spanning from early 80’s till present.
His academic pursuits led him to studies in the USA, where he honed his skills. Sasha
contributed significantly to research and development roles in the defense industries,
accumulating invaluable experience. His expertise in Evolutionary AI spans over four
decades.
He is the main developer of the solution tools of SkyOptim.
Algorithm Specialist & External Collaborator
Karim Mehadhebi works as a Civil Servant in French Civil Aviation. He started as a researcher and has published several innovative methodologies in this field.
He is no longer involved in research as a civil servant, and has agreed with his
administration to work on this field privately, during his leisure time.
His skills in mathematics has helped him to design the specific algorithms used in the lower layers of pattern
identification.