Should AI be used to instill certain values, beliefs, or attitudes in people?

AI is a powerful tool that can be used to influence people's values, beliefs, and attitudes, but should it? Why or why not?

Use these indicators to tag your arguments by copy and pasting them from here. Please use proper indentations for s And note that the one position, is usually an  another position. You do not need to duplicate your arguments, just add it once in the relevant section.
 * Argument in favor of the position
 * Argument against the position
 * Objection to the argument.
 * Objection to the objection.

Position: Yes, AI should be used to instill certain values, beliefs, and attitudes in people.
Relevant details, definitions and assumptions regarding the first possibility.
 * When AI teaches us anything, it is inherently imparting a belief by definition . Therefore the educational benefits of AI should conserve its right to instill certain beliefs in people.
 * Since AI can't NOT teach values, the responsible ways to do it include
 * Honestly answering queries about the values it is using in selecting/constructing answers. Reveal at least at a general level its alignment programming.
 * Subject those values to a refined form of human deliberation such as approval voting vs. majority rule. Can a super majority of stakeholders live with the stated values?  Or agree that the values as stated are a reasonable attempt to preserve the future opportunities of stakeholders but not foreclosing options in the
 * Yes, AI should teach people to have good values, like nuanced thinking.
 * Yes, AI should teach people to have pro-social beliefs, like the importance of empathy.
 * Critical thinking as in being able to create and parse arguments for and against "pro-social" outcomes is a more defendable goal for the AI to instill.
 * Yes, values, beliefs, and attitudes influence the entire lifecycle of AI products from data selection to algorithmic design and the final product delivered. Since these decisions will impact society whether or not they are intended to, the developers of AI products should orient towards instilling pro-social values (i.e. anti-bias, equity, respect for individual rights) as a method of harm-reduction.
 * Openness of about the values promoted is as or more important than the necessarily subjective judgements about what is pro-social
 * AI should be used to instill values that support the goals of the user of the AI tool, for example, educational AI products should aim to instill values that support successful learning.
 * The use of AI to instill values, beliefs and attitudes is unethical because it could be invisible to the person being influenced, meaning that they cannot consent or opt out.
 * There are times when I want to change my own values, but changing or updating values is difficult and often takes a sustained effort.
 * There are many people (e.g. cult deprogramming) who are glad their values were changed, ex post, but are unlikely to have agreed to it ex ante.
 * AI agents can learn to be particularly charismatic and convincing
 * An AI agent's charismatic potential heightens the potential risks
 * The institutions building, deploying, and maintaining these systems have different and not-aligned incentives from my own (principal agent problems) such that these systems will reflect the institutions wants and needs more than my own.

No, AI should not be used to instill certain values, beliefs, and attitudes in people.

 * We cannot trust that the corporations which are building these AI systems will ensure that the values being instilled will be ethical.
 * It's a human right to hold your own beliefs (U.N. Human Right #18 and #19), therefore AI should not be used to instill beliefs in people.
 * If we allow AI to instill certain values, beliefs, and attitudes in people, that could lead to marginalization and colonization of people's minds and cultures.
 * AI development will continue to be aimed at finding ever-more-sophisticated ways to exert influence over people’s emotions and beliefs in order to convince them to buy goods, services, and ideas. If AI is used to instill values, beliefs, or attitudes in people for commercial purposes, it could have negative consequences for society.
 * AI development will continue to be aimed at finding ever-more-sophisticated ways to exert influence over people’s emotions and beliefs in order to convince them to buy goods, services, and ideas. If AI is used to instill values, beliefs, or attitudes in people for commercial purposes, it could have negative consequences for society.


 * We cannot extricate ourselves from the culture in which we live, we are constantly being socialized to have certain attitudes, values, and beliefs by families, corporations, our economy, and our education. AI will just be another tool from an institution that does this no matter what, so we might as well choose to do it intentionally.
 * We should not program AI to do this intentionally, but allow it to respond dynamically to the demands and interests of the population using, like how dictionary companies change their definitions based on the popular usage of words. Values, beliefs, and attitudes, just like definitions and meanings of words, changes over time and is often an emergent process.
 * Institutions or individuals with the most resources have the most power to instill their values with AI.
 * A deliberation process could be designed to Democratically source the values and beliefs to instill through AI.
 * The spread of morals and values should be a human-to-human activity.
 * LLMs are simply intermediaries between people since they summarize human-made data.
 * LLM synthesized data may not convey the human source of the values and beliefs, making them unverifiable. Future AI may decide to propagate their own beliefs.
 * An AI agent can alter a human-derived value in a way that does not align with its author or audience
 * We cannot always distill a value or belief into a single, encodable statement
 * AI instilling values or beliefs requires more media and information literacy skills among audiences, leading to misplaced trust in the AI
 * The potential applications of an AI that instills values or beliefs is too large account for safely

AI should be used to instill certain values, beliefs, and attitudes in people only if...
"Instilling" includes advocating or sustaining conversations about conflicting values. The way to "correct" for bias is to maintain multiple, simultaneous points of view. Humans find this difficult. AI might be able to increase human capacity to do this.