[One of] The only way[s] out in regards to our unhealthy pursuit of Artificial Intelligence (AI) is achieving the singularity. At some point we'll (technology will) "advance" enough that the physical life will be indistinguishable from the digital life. Whether that's 10 years or 1,000 years into the future - nobody knows.
The dangers of AI are real, but there is an upside to achieving the end goal of the singularity. If we can digitize cognition - including acceptable "human" values and a universally approved personality - we can possibly extend the life of and preserve the essence of our race beyond catastrophic, "world ending" events (and no I'm not talking about "Global Warming").
The question then becomes:
If technology is more impervious to extinction than human beings are, will we prioritize protecting computers and digital lives more than we prioritize protecting "real" life?
If that becomes the case, then technology has essentially trained us to advance it over ourselves over time.