it's just an advanced eliza bot with a 40000 word working memory. it doesnt reason, and therefore cant determine the accuracy of a statement. considering it uses a fixed dataset, it's normal to expect factual errors in its interpretation of the data. it knows nothing about the world, and has no sensing apparatus. it's literally just trying to figure out what you want to hear, not drawing any conclusions about it.
you claim to know a lot about AI implicitly when you say that you worry about those who dont know as much, but this article would suggest otherwise, since you clearly dont understand the limitations of a language model. it's not answering your query about drinks, it's pretending to hold a conversation on the topic. it's a fine distinction which you obviously missed.
when you're done anthropomorphizing it and projecting your feelings and biases all over the place, maybe try asking it to solve a math problem, and you should understand the limitations im speaking of. it's interesting that, since code is a linguistic representation of formal logic, it could create a script that solves a problem, but it cant actually think its way through the problem.
So, the machine you interacted with had (HAS) all the intention to dupe, lie, deceive, cause you to stumble on the search for truth...? Is this malice?? Did this machine's acting offend you???
Individuals who gaslight are childish, uncertain, and intellectually harmful. They hope to deprecate individuals to inspire their own self images.
https://www.outlookindia.com/outlook-spotlight/11-best-fiber-supplements-powder-gummies-pills-effective-for-constipation-weight-loss-gut-health-etc-news-287970
it's just an advanced eliza bot with a 40000 word working memory. it doesnt reason, and therefore cant determine the accuracy of a statement. considering it uses a fixed dataset, it's normal to expect factual errors in its interpretation of the data. it knows nothing about the world, and has no sensing apparatus. it's literally just trying to figure out what you want to hear, not drawing any conclusions about it.
you claim to know a lot about AI implicitly when you say that you worry about those who dont know as much, but this article would suggest otherwise, since you clearly dont understand the limitations of a language model. it's not answering your query about drinks, it's pretending to hold a conversation on the topic. it's a fine distinction which you obviously missed.
when you're done anthropomorphizing it and projecting your feelings and biases all over the place, maybe try asking it to solve a math problem, and you should understand the limitations im speaking of. it's interesting that, since code is a linguistic representation of formal logic, it could create a script that solves a problem, but it cant actually think its way through the problem.
It should have been using a different term, like “perceived sweetness.”
So, the machine you interacted with had (HAS) all the intention to dupe, lie, deceive, cause you to stumble on the search for truth...? Is this malice?? Did this machine's acting offend you???
Be well.
H.Rojas