There are a host of technologies currently relevant to the development of robust machine intelligence. Without going into too much detail on them:
Cognitive Neuroscience has been revolutionizing our understanding of the brain and it's processes. Since the brain is, at present, our best (and perhaps only) example of an intelligent, conscious system, future learnings in this area will prove invaluable.
Computing disciplines are obviously front-and-center. "Top-down" AI focuses on language aquisition, rule-based systems, and prescriptive behavioral response systems. "Bottom-up" AI (Neural Nets) targets learning and pattern recognition. These two approaches were split into rival camps for many years (the "perceptron" controversy), but each has valuable uses. Future developments will almost certainly rely on findings from both sides. I would be remiss here if I failed to mention the impact of new hardware such as VLSIs (Very Large System Integration), FPGAs (Field Programmable Gate Arrays), and CAMs (Content Addressible Memory).
Computational Techniques that will play a role in modelling learning and intelligent, goal-directed behavior include Mathematical Optimization, Stochastic Processes, Decision Theory (Beyesian and otherwise), Game Theory, and Dynamic Programming.
"Esoteric" disciplines like Chaos Theory, Artificial Life,
and Emergent Systems also speak to these issues.
The future history of consciousness