You can read the details here. https://www.axios.com/2024/06/24/chat-gpt-generative-ai-perplexity-hallucinations
One of the AI tools the article talks about is Perplexity. I meant to write about this earlier but I forgot about it. I tried it as an AI search engine. I thought it was terrible. Just for chuckles I thought to ask what is the reputation of my website chimac.net. It said that the website wasn’t trustworthy for a variety of reasons including that it talks about scams.
Now this is not the option that other website review trust sites have. They note how long I have been existing, and that I don’t have any known malware or other issues on websites. Everything that Perplexity said was BS about my site, and I responded to it and reported it and identified that I was the owner of this site and why their reasoning was incorrect. It was just remarkable. There was nothing AI in that response. It didn’t gather any facts, it just said an opinion that had no basis.
Earlier on LinkedIn someone was positive about AI and I shared the link above and asked about the problem of AI making stuff up. He didn’t respond to it. It is one thing to make a claim, but when our claims are as baseless as an AI what exactly is the difference between us and an AI? Nothing on LinkedIn.
One of the reasons that I don’t use AI is that I don’t see the value when you know more than the answer. I am not saying I know everything. It is just that when I look for an answer I have a range of what a likely answer will be, and when it gives me something completely unexpected then a little more digging almost 100% proves the AI is wrong. It doesn’t matter if it’s ChatGPT or whatever. AI sounds good when you are stupid and don’t know any facts or are too lazy to do research yourself.
Do you remember the movie Wall-E in which the future humans floated on chairs and were served by AI? We are becoming them faster than we think, and it will be our downfall to give up critical thinking to a machine.