Ashby assumes a unique position in the history of cybernetics. Alongside N. Wiener, J. Bigelow, and C. Shannon, he helped create the academic trend. With his contemporaries, he shared the goal of artificially reproducing complex human behavior by exploring the common ground of machinery and organisms. However, Ashby’s approach in his demonstration at the 1952 Macy Conference was so shocking to the point that it caused the established cybernetics models to crack. Because of such novelty it was difficult for his colleagues to accept his view. At the core of the discussion was the Homeostat. With this device, Ashby integrated all his academic hypotheses on the interactions between the nervous system and the environment. Since the method in which this mechanical brain adapted itself to and learned the environment had already predicted the wave of the next generation cybernetics, his research had both ambiguous and ambivalent aspects. Focusing on Ashby’s Homeostat, this article will approach the integrated essence of the biological process and cognition into which he gained insight, using the notions of ‘random mechanism’ and ‘ultrastability’ that formed the basis of this device. These two core elements not only clearly highlighted the difference between the established cybernetics models and the Homeostat, but also embraced the methodological detour of cognitive science and artificial intelligence afterward. Accordingly, this article attempts to relocate Ashby’s position, typically described as a founding member of cybernetics, as the instigator of second-order cybernetics and the originator of the recent discussions in the biology of cognition and artificial life.