Raw LLM Responses

My grandfather in the Philippines suffered from a serious illness that affected his heart and lungs, and despite seeing multiple doctors, we struggled to get clear answers or a treatment plan that made sense. Out of desperation, I uploaded his lab results to ChatGPT, not for a final diagnosis but for another perspective. It estimated that he might have less than a year to live. He passed away seven months later. That experience didn’t make me think “AI replaces doctors.” What it did show me is how badly human systems can fail. In the Philippines, we had a very hard time finding doctors who would actually listen to us, explain things clearly, or honor the tests and requests we were making. Many of the doctors and nurses we dealt with were condescending, hiding behind their titles and education, and treating us like we were ignorant for asking questions about my grandfather’s care. It often felt like we were being talked down to instead of partnered with. My grandfather eventually told us he was ready to stop fighting because he was exhausted by the constant procedures, conflicting opinions, and feeling like nobody was really helping him, just poking him with needles and guessing. I did my best to push for what I thought he needed, but it’s hard for families to be direct and assertive in a culture where questioning medical professionals is seen as disrespectful. So when people blame ChatGPT for someone’s poor choices, I push back. ChatGPT didn’t fail my grandfather; the humans and the healthcare system around him did. AI is a tool. It can be helpful and offer perspective, but it doesn’t make decisions for us. Blaming AI for human misuse only distracts from the real problems: Doctors who don’t listen, nurses who dismiss families, hierarchical cultures where patients are afraid to speak up, and professionals who don’t take seriously the people who are fighting for someone they love. I don’t blame ChatGPT at all. I blame the incompetent humans and broken systems that failed my grandfather at every level.