Because autopilots can be engaged and disengaged at will, we can cruise along trusting in the course steered by the principle of social proof until we recognize that inaccurate data are being used. Then we can take the controls, make the necessary correction for the misinformation, and reset the autopilot.
