Yeah I’m quite pleased that these AI models are just like any early version of lots of software. Unintended consequences turn into known problems when you use the chat bot for support.
I started to write a list of the complete bollox sometimes produced by Perplexity. I gave up when it became apparent the task was Sisyphean. Perplexity, and I guess other web search engines that use AI, is very effective in providing quick answers compiled from different sources, with links to those sources. But—a big BUT—you have to treat those answers with a huge degree of scepticism. As you say Jack, AI wants to make you happy by giving you answers. It cares not about the accuracy of those answers. (I would make a comment about my time living in South-East Asia, but it would come out sounding racist.)
Am I to believe that there are people seeking answers from AI? That is not what AI is for and was never intended to be for. AI can help you find an investment advisor if you wish to help a grandchild starting a job and who wants to save for the deposit on a house. It can tell you that Chinese cars use Chinese as their source language for their user manual so don’t expect to be told where the light switch is, as you would if it was an English car. My father was not a speaker so I never got any answers from him, but that encouraged me to advise my children on the options available to them when I had the opportunity. I made it to University when I was 23, my son made it when he was 27 and had already created two businesses, my daughter by the age of 25, having decided to become a professional photographer. None of these options would have been recommended by AI. As for Nazi bias, what on earth is such a system doing on the internet in the first place, and how can it remain there after it is discovered?
Yeah I’m quite pleased that these AI models are just like any early version of lots of software. Unintended consequences turn into known problems when you use the chat bot for support.
I started to write a list of the complete bollox sometimes produced by Perplexity. I gave up when it became apparent the task was Sisyphean. Perplexity, and I guess other web search engines that use AI, is very effective in providing quick answers compiled from different sources, with links to those sources. But—a big BUT—you have to treat those answers with a huge degree of scepticism. As you say Jack, AI wants to make you happy by giving you answers. It cares not about the accuracy of those answers. (I would make a comment about my time living in South-East Asia, but it would come out sounding racist.)
Am I to believe that there are people seeking answers from AI? That is not what AI is for and was never intended to be for. AI can help you find an investment advisor if you wish to help a grandchild starting a job and who wants to save for the deposit on a house. It can tell you that Chinese cars use Chinese as their source language for their user manual so don’t expect to be told where the light switch is, as you would if it was an English car. My father was not a speaker so I never got any answers from him, but that encouraged me to advise my children on the options available to them when I had the opportunity. I made it to University when I was 23, my son made it when he was 27 and had already created two businesses, my daughter by the age of 25, having decided to become a professional photographer. None of these options would have been recommended by AI. As for Nazi bias, what on earth is such a system doing on the internet in the first place, and how can it remain there after it is discovered?