Do you see the possibility of making Computer Science PhD work as a some real product rather than just few publications?