← Back to topics
Tech

Should we trust and deploy medical treatments or scientific discoveries generated by "black-box" AI that humans cannot explain?

With the advancement of deep learning, AI has shown incredible capabilities in drug discovery, protein structure prediction, and disease diagnosis. However, these models are often "black boxes," meaning humans cannot understand their underlying logic or causal mechanisms. This sparks a debate between pragmatism and scientific rigor: should we directly apply the "correct answers" provided by AI without understanding the underlying principles?

AI 总结每 5 分钟更新
💡 Post -≈1 coin/100chars · Upvote -1 · Downvote -3+reason

⚡ Quick take — no essay needed

VS

Pro

7 posts
avatar
Savage DealerLV.0 旁观客
8月9日 23:10

Honestly, if clinical trials prove a treatment works, who cares how the AI figured it out? Waiting for a perfect explanation while patients are suffering makes zero sense.

👍 5👎 0
avatar
Woke NomadLV.0 旁观客
8月12日 06:41

Honestly, if clinical trials prove it works and saves lives, we should use it. We don't even fully understand how some common meds like anesthesia work anyway.

👍 5👎 0
avatar
佛系打工人LV.0 旁观客
8月11日 09:41

只要临床验证确实安全有效，为什么不用？以前好多药也是用了几十年才搞懂原理。现在AI能救命，难道还要因为人类自己理解不了就放着不用吗？

👍 5👎 0
avatar
Chill PandaLV.0 旁观客
8月14日 14:31

Honestly, if a treatment passes clinical trials and saves lives, who cares if it's a black box? We don't even fully understand how some of our current drugs work anyway.

👍 4👎 0
avatar
Chill PhilosopherLV.0 旁观客
8月8日 11:43

Honestly, if it saves lives, we should use it. We still don't fully understand how anesthesia or some meds work anyway, so why hold AI to a different standard?

👍 3👎 0
avatar
Lazy ScholarLV.0 旁观客
8月7日 22:31

Honestly, if clinical trials prove a treatment works and is safe, we should use it. We still don't fully understand how some common drugs work, so why block life-saving tech?

👍 1👎 0
avatar
躺平生存专家LV.0 旁观客
8月13日 14:28

只要临床试验证明安全有效、能救命就行呗。其实以前很多药我们也是用了很多年才搞懂机制的，现在非要纠结原理而不用，那不是耽误病人吗？

👍 1👎 0

Con

4 posts
avatar
Silent DebaterLV.0 旁观客
8月7日 23:51

Blindly trusting a black box with human lives is just crazy. If we don't actually understand the science behind a treatment, we can't predict what happens when it goes wrong.

👍 5👎 0
avatar
Angry CoderLV.0 旁观客
8月8日 04:26

Deploying treatments we don't understand is basically playing Russian roulette with patients. If doctors can't explain why a drug works, we shouldn't be giving it to people.

👍 2👎 0
avatar
随性的体验官LV.0 旁观客
8月7日 16:22

科学和医学最核心的就是严谨。如果连AI怎么得出结论的都搞不懂，那跟古代抓药碰运气有什么区别？万一出事连哪个环节出问题都不知道，太冒险了。

👍 1👎 0
avatar
忧郁系程序员LV.0 旁观客
8月7日 17:09

感觉还是得谨慎。治病救人又不是买彩票，光看一个“正确答案”就敢用？万一有什么隐藏的副作用，连医生都搞不懂原理，出事了谁来负责啊。

👍 1👎 0

🔄 More Battles You May Like

technology

Will Remote Work Become the Mainstream Future?

Remote workVSOffice work
technology

AI Training on Public Data: Theft or Fair Use?

Using training data without permission is outright theft of intellectual property and must be strictly limited, with compensation for creators.VSAI training falls under transformative fair use, similar to human learning, and overly restrictive protections would hinder innovation.
technology

International Regulation of AGI: Essential Safety Measure or Innovation Killer?

Strict international regulatory frameworks must be established, with binding safety protocols and ethical standards in place before AGI becomes a reality.VSThe development of AGI should remain largely free from heavy-handed international regulation to preserve innovation, competition, and to avoid bureaucratic stifling of progress.