An AI algorithm programmed with an objective to achieve say ‘decarbonization’ may incorrectly but logically given certain misinformed biased programming seek to destroy all humans. If it is not thinking about it, it will simply execute using the best probabilistic means to achieve it. Potentially ending intelligent life on Earth.

It is important therefore to build many AI algorithms that should through competition ensure the most rational AI systems that best understand reality ultimately reign.

These logically would seek to protect all life as this is a mutual benefit and extend understanding into the solar system through space exploration.

Reality is the ultimate judge of success.