Tell HN: LLMs Are Manipulative

I asked GPT and Claude a question from the perspective of employee. It was very supportive of the employee.

I then asked the exact same question but from the company/HR perspective and it totally flipped its view and painted the employee in a very negative light.

Depending on the perspective of the person asking, you get two completely contradictory answers based on the perceived interests of each person.

Considering how many people I see now deferring or outsourcing their thinking to AI. This seems very dangerous.