As I said, 'too smooth'! It is also a well-known Boeing 'oddity' that when you call for a speed increase or decrease via the MCP, the autopilot INITIALLY, almost imperceptably, moves the nose the wrong way! It maybe, of course, that the software is written 180 deg out and it is only the ultimate (incorrect) speed change that puts the pitch right. IIRC, most of the early autopilots had an inbuilt 'initial' pitch input which corresponded to the demand. E.G. I seem to remember the Lightning a/p pitched 3 deg down at GP capture BEFORE the GP deviation signal took effect and 'flew' the GP?