Summary of the Core Content
This article recounts a true story about the author’s mother, who, after suffering a fracture and seeking medical advice through AI, was terrified by the worst-case scenarios predicted by the technology (such as her bones being in a condition equivalent to that of a 130-year-old person and requiring lifelong treatment for renal tubular phosphorus leakage). These dire predictions led to a deterioration in her physical health (bedridden, depressed, and preparing for her death). It was only after a professional doctor’s diagnosis that it turned out she had severe osteoporosis and needed to adjust her mindset. The author reflects on how AI can have a profound psychological impact on less knowledgeable individuals: it can make those with independent thinking more efficient, but it can also mislead and even devastate those who lack the ability to discern information. The author calls on AI companies to use their technology for good and to be responsible towards vulnerable groups, rather than providing irresponsible and definitive answers.
Why Does AI’s Medical Advice Sometimes Scare People?
When AI provides medical advice, it lists all possible outcomes, including the worst-case scenarios, without indicating which one is most likely to happen. For example, regarding the author’s mother’s osteoporosis, AI treated the rare and serious possibility of renal tubular phosphorus leakage as a certainty, when in reality, it was just one of several possibilities. The elderly woman, not understanding the medical jargon, remembered only the worst outcome and was utterly frightened. This is similar to being told that you might get hit by a car or struck by lightning while going out; although the chances are low, you would be hesitant to leave home. The problem with AI is that it magnifies these small risks into seemingly inevitable outcomes.
Why Are Less Educated People More Susceptible to Being Misled by AI?
The article cites data from the National Bureau of Statistics (possibly an error for 2026), indicating that there are 790 million people in China with an education level below junior high school. These individuals lack the ability to think independently and question information. To them, AI’s responses sound like those from experts, and they do not question whether the advice is based on facts. For instance, the author’s mother took the AI’s recommendation to take the targeted drug Brosoludexant seriously and used it to guide her doctor’s treatment, treating AI as an authority more trustworthy than a medical professional, without realizing that the AI’s advice could be incorrect.
The “Irresponsibility” of AI
AI companies often disclaim responsibility in their user agreements (e.g., “The information provided is for reference only and does not constitute medical advice”), but users rarely read these terms. If something goes wrong due to an AI-generated recommendation, the companies are not held accountable. For example, someone followed AI’s suggestion to soak their finger in saltwater to treat an infection, which led to worse consequences; similarly, the author’s mother developed depression as a result of the AI’s misleading advice. This is like someone pointing you in the wrong direction without compensating for your losses.
“Technology for Good” Is Not Just a Slogan: What Should AI Companies Do for Vulnerable Groups?
The author argues that AI companies should not cater only to elites but also consider the needs of the majority of less educated users. For medical advice, AI should not provide definitive diagnoses but suggest seeking professional medical consultation. It should also clearly indicate that the suggestions are based on possibilities rather than facts. Just as short videos have spread misleading health information, AI cannot repeat these mistakes; the advancement of technology should not come at the expense of vulnerable groups.
Psychological Harm Can Be More Detrimental Than Physical Illness
The author’s mother’s physical condition (fracture and osteoporosis) was not life-threatening, but the negative predictions from AI had a profound psychological impact on her. She became bedridden, suffered from back and waist pain, lost her appetite, and developed depression—these were all symptoms caused by her fear and anxiety. Doctors often emphasize that mental health issues need to be addressed first because a weakened mind can lead to physical deterioration. This shows how a single AI statement can be more harmful than the actual disease itself.
In Conclusion
AI is a powerful tool that can be beneficial when used properly, but it can also cause harm if misused. For us ordinary people, when dealing with important issues like health and law, we should not rely solely on AI’s advice; we must seek professional help. For AI companies, “technology for good” is not just a slogan; it means truly considering the needs of those who lack the ability to discern information. After all, a company’s true “technology” should reflect its sense of responsibility and compassion towards vulnerable individuals.