There is a lot of research done right now, funded by the EU. But uncertainties are huge, and the tipping point for the Amoc could be between 2050 and 8065 [0].
It's a complex problem, and almost impossible to model correctly, but that doesn't make it science fiction.