Both recommender systems and bots should proactively and smartly answer the questions of software developers or other project stakeholders to assist them in performing their tasks more efficiently. This paper reflects on the achievements from the more mature area of Recommendation Systems in Software Engineering (RSSE) as well as the rising area of Bots in Software Engineering (BotSE). We discuss the similarities and differences, briefly review current state of the art, and highlight three particular areas, in which the full potential is yet to be tapped: a more socio-technical context awareness, assisting knowledge sharing in addition to knowledge access, as well as covering repetitive or stimulative scenarios related to requirements and user-developer interaction.