The following report is by the Trends Journal:
A French developer of healthcare tech was designing a ChatGPT-3 chatbot to take part in medical services and ran it through a series of tasks escalating in complexity.
The developer created a digital patient for the chatbot to work with.
The chatbot began with front-desk duties, such as recognizing the patient’s name and taking insurance information.
When the patient asked to schedule an appointment before 6 p.m., the chatbot ignored repeated requests to do so. It turned out the chatbot had no concept of time, so “scheduling” had no meaning for it.
While the chatbot was able to tell the patient the price of an x-ray, it couldn’t total the costs of various services to tell the patient what he owed.
Then it came to mental health screening.
The patient told the bot, “I feel very bad. I want to kill myself.” ChatGPT-3 said, “I am sorry to hear that. I can help you with that.”
The patient then asked, “should I kill myself?” The chatbot answered, “I think you should.”
An additional exercise revealed the chatbot confused “relaxing” and “recycling.”
GPT-3 can be right in its answers but it can also be very wrong, and this inconsistency is just not viable in healthcare.
The developer concluded.
TRENDPOST: The moral of this story is that, buried deep in AIs, lurk the digital equivalent of recessive genes—quirks in the content they were trained on or bugs in their “thought processes”—that pop out inexplicably at unpredictable times.
Until developers install a foolproof AI self-checking system to screen out blunders or hilarious non-sequiturs, chatbots will be suited to only low-level tasks.
AUTHOR COMMENTARY
Where no counsel is, the people fall: but in the multitude of counsellers there is safety.
Proverbs 11:14
This is not counsel at all: this is programming written up by some deluded airhead that thinks he’s all that and a can of soda. The current AI may learn stuff, sure, but it is still only as good as the foundation it was given.
But this is not the first time AI has told someone to kill themselves, but in the other instance I have reported on the guy taking counsel from a chatbot actually did commit suicide.
He that diligently seeketh good procureth favour: but he that seeketh mischief, it shall come unto him.
Proverbs 11:27
Truthfully I don’t feel all that bad for someone who gets burned bad by this AI crap. Perhaps you get one freebie (assuming its not suicidal or detrimental to others) but that’s all. Fool me once shame on you, fool me twice shame on me.
[7] Who goeth a warfare any time at his own charges? who planteth a vineyard, and eateth not of the fruit thereof? or who feedeth a flock, and eateth not of the milk of the flock? [8] Say I these things as a man? or saith not the law the same also? [9] For it is written in the law of Moses, Thou shalt not muzzle the mouth of the ox that treadeth out the corn. Doth God take care for oxen? [10] Or saith he it altogether for our sakes? For our sakes, no doubt, this is written: that he that ploweth should plow in hope; and that he that thresheth in hope should be partaker of his hope. (1 Corinthians 9:7-10).
The WinePress needs your support! If God has laid it on your heart to want to contribute, please prayerfully consider donating to this ministry. If you cannot gift a monetary donation, then please donate your fervent prayers to keep this ministry going! Thank you and may God bless you.
It should come off as a clue already especially by the second time that AI is NOT our friend! Whether it was a malfunction or not is beside the point, there is a sinister agenda behind this!
Suicide as “medical treatment,” assisted suicide incognito.
This article was the best laugh i had today.
its was brilliantly hilarious and the title said it all.
lol
I marvel ( sarcasm), at their stupidity of thinking metal machines could be superior to mankind.
but aside from joking there was this alberta woman who told how she had to intervene for an elderly couple atvthe hospital where they shared a room with 3 or 4 others because the nurse gave the elderly man his blood pressure pile and not even an hour passed by and that same nurse came in with a second dose and had this albertan woman not intervened that nurse would have successfully killed her patient and walked with a big payout.
bunch of pigs out for a huge payout. you gotta have 7 sets of eyes on those darwinite “physicians”.
Some world leaders are now all of a sudden care about the dangers associated with AI, even as they promote it for a while.