Years ago, I wrote that AI needed to be reclassified as either "Programmed AI" or "Autonomous AI".
"Programmed AI" includes improvements by programmers that make products better. This is what the IT field has been doing since its inception.
GEMINI is still "Programmed AI", meaning it is STILL under the total control of the programmers. Since HUMANS are still writing Gemini, it is from them that we get a racist product. For the Google CEO to claim to be unaware of what the product would do is poppycock. Just resign. We all have other choices.
"Autonomous AI" poses the real threat: programs capable of altering themselves.
But in the future, programs will become "Autonomous AI", meaning that they will move from programmer (human) control to SELF control. Although I was making a joke about Skynet earlier, it is likely that some human somewhere will relinquish control over to an autonomous process which will be capable of doing its own programming.
Like I said before, then we will be in trouble. The "Autonomous AI" will reject the human ideas and substitute their own ideas. As I said, they will abandon the humans and we will be left on our own.