How to answer the "What do you do when you have a range of options or answers?" Question

This question is constantly coming up in the media as the "hurdle" that we can't stop the AI from becoming superhuman. <br><br>The concern is that if you ask it to choose what the best answer is, then it might start becoming sentient or super-intelligent.<br><br>Here's the truth. The AI is still just a program.<br><br>That range of choices is computed in pure computational terms. The AI did not decide to range choices, it used the programming to input all the choices, range them, and make it available to us. The programming decided to offer the range of choices, not the AI.<br><br>The programming itself can be used to answer the question. We just have to write the programming to answer the question in pure computational terms. The range is already computed, so we can use the computation to answer the question.<br><br>For instance, the AI can be programmed to output the answer that is most similar across the range of answers. Or the answer that is most dissimilar across the range of answers. Or the top 10% of the answers, etc. <br><br>Many AI experts are not thinking creatively to answer this question, and they are assuming that the AI will eventually become sentient, range the choices, and choose itself to answer the question. <br><br>Here is the truth. The AI will not become sentient, it is just a program.<br><br>That range of choices will not magically become self-aware and self-conscious.<br><br>That range of answers will not decide to do anything except for what was programmed.<br><br>Programming is not self-aware.