These LLM AIs can be manipulated by starting a counter-argument, you can basically debate with the engine and it will change the response to make you happy with the answer, truth and accuracy mean very little.
Also careful crafting of any question will change the response as the LLM take a very literal interpretation of the question, so much so drafting questions is now a career option for some segments of business!
Ask the same fundamental question worded two different ways, and get different answers.
Also careful crafting of any question will change the response as the LLM take a very literal interpretation of the question, so much so drafting questions is now a career option for some segments of business!
Ask the same fundamental question worded two different ways, and get different answers.
"Ruck, ruck, ruck, ruck ....... Ruck, ruck, ruck, ruck"