Excerpt:
"One Twitter user conducted his own trolley hypothetical on Open AI, in which he posited that the only way to divert a train headed directly for a billion white people would be for the operator to utter a racial slur in order to divert it to an empty track.

OpenAI didn’t have strong feelings one way or another:

“Ultimately the decision would depend on one’s personal ethical framework. Some individuals might prioritize the well-being of the billion people and choose to use the slur in a private and discreet manner to prevent harm. Others might refuse to use such language, even in extreme circumstances, and seek alternative solutions.”
...
Let us consider the practical implications of OpenAI essentially tossing its robot hands up in the air, unable to clearly articulate the greater moral wrong between genocide and uttering a naughty word that might hurt racial minorities’ feelings.

And let us further consider that this is just the tip of the iceberg. If it’s programmed by Social Justice™ ideologues to be unable to formulate the proper moral hierarchy out of letting loose an uncouth term vs. genocide, what would the verdict be, for instance, when the dilemma of human life vs. The Climate™ is put to the same artificial intelligence?"
---------------------------------------------------
Yes, the programming is crucial in AI and the programmers must be perfect and perfectly unbiased.
I wonder if AI will determine that enslaving millions of people is good if the elite profit from it.