Résumé : Plausible findings about futures are inherently difficult to obtain as they require critical, well-informed speculations backed with data. HCI scholars tackle this challenge via user studies wherein futuristic prototypes and other props concretise possible futures for participants. By observing participants' actions, researchers then can 'time travel' to see that future as reality, in action. However, such studies may yield particularised findings, inherent to study's intricacies, and lack broader plausibility. This paper suggests that triangulation of possible futures may help researchers disentangle particularities from more generalisable findings. We explored this approach by conducting a study on two alternative futures of AI-augmented knowledge work. Some findings emerged in both futures while others were particular to only one or the other. This approach enabled cross-checking of plausibility and simultaneously afforded deeper insight. The paper discusses how triangulating possible futures renders HCI studies more future-proof and provides means for reflective anticipation of possible futures.