ARTIFICIALLY intelligence chatbot ChatGPT is causing controversy at the moment over its creepy ability to write just like a human.
Students are using it to cheat on exams, copywriters are fearing for their jobs and some people are even using its ‘advice’ as a replacement for dentists and counselors.
AI expert Lisa Palmer spoke exclusively to The U.S. Sun about ChatGPT concerns and the jobs she wouldn’t trust it to do right now.
Palmer has largely positive views on the AI chatbot and thinks we all should be using it.
However, she thinks we need to be wary of its limitations right now and not put pure faith in it to do human jobs.
The AI expert told The U.S. Sun: “I think that there is a replacement of tasks happening for jobs.
“We are seeing [human] tasks being replaced by artificial intelligence.
“It’s not very frequent to see an entire role replaced by AI in general.
“ChatGPT specifically, I certainly would not trust it at this point to replace the work that I do.
“If you have tested it you see that it makes things up that are quite nonfactual from time to time.”
Speaking on reports of people in Hong Kong and China using ChatGPT for medical and dentistry advice, Palmer urged caution.
She warned: “I definitely wouldn’t recommend that.
“The very best applications of artificial intelligence are when we have human plus machine so if we want to get the best outcomes we have to think about creative ways for humans to use artificial intelligence as a tool.”
The expert added: “Just outsourcing entire functions, particularly things like a doctor, that would be ill-advised at a minimum.
“You don’t want to outsource that kind of expertise to any kind of AI, ChatGPT or otherwise.”
Palmer also raised concerns about using AI as a replacement for a lawyer.
A company called DoNotPay, has developed what it describes as “The World’s First Robot Lawyer”.
That AI robot lawyer will soon be defending a human in court.
Paler told The U.S. Sun: “In the story about them rolling it out for a lay person to take advice through an earbud, I think is a bit dicey with regard to how that will turn out.
“However, if we were to put that same tool in the hands of a lawyer and be able to feed them with on-point case law in the middle of an ongoing debate in trial, I think it could be very powerful.
“So to me, it’s about making sure that we are thinking about using ChatGPT as an augmentation for humans that have the knowledge to know when it’s right, when it’s not right and when it’s introducing risk that the AI doesn’t know.”