Indeed, I think we're going to be seeing a lot of task-specific AIs before a truly generalized AI shows up.
You may be interested in
OpenCog, from the project website:
No challenge today is more important than creating beneficial artificial general intelligence (AGI), with broad capabilities at the human level and ultimately beyond. OpenCog is an open-source software initiative aimed at directly confronting that challenge.
The intend to have a system that can do artificial learning and conduct biology experiments in a few years (2017-2018)
Creation of an OpenCog-based artificial scientist, operating a small molecular biology laboratory on its own, designing its own experiments and operating the equipment and analyzing the results and describing them in English.
Full on human level AI in 2019-2021.
Then after that a version that can self-improve (as in work on it's own code) (2021-2023)
Also note that this project is not vapourware they have actual software already and it can do stuff, it also has real AI experts working on it, so it actually has some chance of achieving it's goals (On schedule or not is another thing entirely)
Edited to fix a typo