I’ve just been passed an article from Natural News, reporting how a man nearly killed himself after following what he thought was diet advice from ChatGPT. The answer, apparently, was replacing salt with sodium bromide—a sedative pulled from the human pharmacopeia a century ago for being toxic. He read. He believed. He took. He found himself in hospital with the kind of chemical injury you don’t usually get outside of Victorian suicide notes. He survived, which is a shame.
Now, I’ll freely admit that the availability of medical information from AI engines is one of the best things to happen in the modern world. You no longer need to ring up the GP surgery, then wait on hold behind half the population of Bangladesh, and explain your symptoms to a doctor who may be too too lazy, or too dim to know what’s wrong. You no longer need to sit in a fluorescent-lit waiting room beside wheezing wrecks, staring at posters about chlamydia and colonoscopies. You can just ask a machine. That machine has, or can be given, access to the accumulated medical knowledge of the past few centuries. It is never impatient, never judgemental. It does not lecture you about your decision not to accept a vaccine. It simply answers.
For the intelligent, this is utopia. The clever man uses the tool, compares sources, checks the chemistry, and makes better choices for himself. He asks a question about blood pressure medication, or uploads a picture of the lump on his groin, and gets a detailed reply. He, cross-references it and comes away knowing more than most junior doctors.
But here is the trouble—rather, here is the joy. The same tool is also available to the other kind of man: the man who should not be trusted with an electric kettle; the man who will ask a question to which the answer is “sodium bromide” and think, Yes, I’ll have a bit of that— never asking whether sodium bromide is safe or sane, but swallowing it and waiting to see what happens. In this case, what happened was bromide poisoning. That is sad only if you believe that all life is equally valuable. I don’t.
No reasonable man would have got into this mess. No reasonable man would have thought “sodium bromide” was a healthy addition to his weight-loss programme. And no reasonable man would have started dosing himself without ten minutes’ investigation. The only pity of the present case is that he lived to be interviewed.
Indeed, this is the undiscussed upside of modern technology. Every supposed “risk” has a silver lining. Two years ago, one of the boys at my school managed to kill himself after watching a series of TikTok videos on some moronic stunt about holding your breath. Since he may have pegged out on school premises, the whole thing was covered up. But enough got out, and the teachers couldn’t resist the urge to put on sad faces and offer grief counselling to everyone in sight. I didn’t take it. The boy was fat and smelly. The hair on his head was uncut and greasy, on his legs thick in some places, bald in others. The manner of his departure is all you need to know about his intelligence. Therefore, his departure raised the average IQ in my town and marginally improved the visual environment.
This is why I welcome AI and all its alleged dangers. The clever will use it to get cleverer. The stupid will use it to get dead. Either way, the future improves. You can clutch your pearls about “misinformation” and “safety risks” if you like. I’ll be over here, healthy, informed, enjoying a world where the surplus fool has every opportunity to remove himself from it.
Discover more from The Libertarian Alliance
Subscribe to get the latest posts sent to your email.
“ He survived, which is a shame.”
How edgy.
It would be good if Bryan’s next essay were to address the question: “If an AI gives advice that proves to have been bad, who should be held responsible for the effects?”
Since AI is largely killing off the ugly and generally unfit, Bryan doubtless thinks your question irrelevant. He would rather celebrate the cull.