Future proof: Solving the ‘adaptability paradox’ for the long term