The argument about whether capability increases gradually enough for society to adapt and correct course, or fast enough that decisions made beforehand are all you get. Almost every disagreement about AI policy traces back to a difference of view on this, usually unstated.
Technical view
Fast takeoff scenarios posit a discontinuous jump driven by recursive improvement; slow or continuous scenarios posit steady, broadly distributed progress with economic effects arriving before transformative capability. Evidence to date favours continuity in aggregate, complicated by discontinuities in specific capabilities and by long lags between capability and deployment. The governance implication is which mechanisms can work: iterative regulation requires time to observe and respond.
Related
Part of AI Terminology Index, section: Alignment and Safety Concepts.