There is an amazing quote which I do not remember the author “clear path does not necessarily mean a short path” AGI will not happen in 15 years for sure, might not happen at all in this century, might not even happen at all. As ML/AI researchers we did not show any generalization yet, all our advancements are about machines performing specific tasks.
About loosing jobs, it might be contradictory but automation always created jobs. For example, there are more bank tellers after the invention of ATMs, there are also more factory workers in the world then pre-industrial revolution times. What happens with automation is increase in the efficiency hence workers become more efficient in their jobs and perform them more efficiently, at the end per service/product cost decreases and demand increases. So, the loss of job will probably not be an issue either. I think best reference would be John Markoff’s book for a better explanation.
Although I created more optimistic scenario, there is a huge problem with automation and it is not loss of jobs, it is income inequality. In non-automated world, each employee is responsible for greater part of the service/product than the automated world. Hence, in an automated world when the income is distributed back, it largely goes to the person with a capital since he/she owns the robots henceforth own their product. I think the real issue is somehow solving this and it largely a government issue.