These groups, and the people within them such as Ben Goertzel and Eliezer Yudkowsky, have long been working in the field of Artificial Intelligence and theorizing about how to ensure AI remains “friendly” to human beings, among other ethical considerations. They are amazingly smart; however, they cannot be expected to do this alone, nor should be trust them to. All who are able have a duty to scrutinize this project thoroughly, considering the vast potential consequences. Thankfully, the open source nature of this project will allow it to be scrutinized.
If it is possible, someone, somewhere, will eventually create a recursively self-improving artificial intelligence that will quickly surpass human intelligence. This power could be harnessed for very positive uses, and hopefully it will be a project like OpenCog that does it first, rather than DARPA.