Microsoft Research has developed a learning system that can be used by lifts to understand where individuals need to go.
Without using facial recognition software, the system is able to monitor people's movements and history through sensors outside the lift in order to learn which floor they intend to travel to.
"Without any programming, we just had an AI system that just watched what people did and did that for about three months," Peter Lee, head of Microsoft Research, told Bloomberg.
"Over that three months the AI system started to learn 'this is how people behave when they want to enter an elevator', and 'this is the type of person that wants to go to the third floor as opposed to the fourth floor'.
"After that training period we switched off the learning period and said 'OK, go ahead and control the elevators'. And so without any programming at all, this intelligent system was able to understand people's intentions and act on people's behalf."
For the moment, the project is still only a concept, with no plans to roll it out on a commercial scale. Significantly, though, the smart lift is part of a larger project aimed at developing Artificial intelligence, currently Microsoft Research's largest investment area.
"There's a general theme we have here in the labs where today people talk about operating a computer," Lee added. "But we think in the future you won't be operating computers but instead computers will be working on your behalf, and this elevator project is just one expression of that idea."