If you ask Google AI which country has the highest obesity rates it will say it is Nauru, Tonga, and Samoa which is correct but then it also adds another paragraph saying that even if these countries have the highest obesity rates . China, India and America have the highest overall numbers of obese people(no fucking shit they would when they have larger population numbers)..but the AI still gotta add that in to avoid the objective reality that those island countries are the fattest countries.
And if you say " Japan has the lowest rates of obesity among first world countries" Google AI says that it is not true that Japan has the lowest obesity rates among first world countries and it says it's because South Korea, Nepal and Ethiopia also have low obesity rates.
Except South Korea still has higher rates of obesity than Japan. While Ethiopia and Nepal's are not first world countries. So the original statement is in fact still true but Google AI claimed that it is false. .....By listing off countries that weren't first world or are still fatter than Japan
In part your frustration is because you don't understand how AI works. These are not logical reasoning engines. You can create two chats and ask it the same question and get different answers. I did this with something about graphics cards and two different sessions, without memory in play, one said the RTX 5090 was the most powerful graphics card on the market today, and the second said it was the RTX 4090.
Interjecting wokeness aside with "well akschually it's because..." is one thing. But getting curiously 'correct but not' answers is because of how LLMs work. They used to make everything up. They're improving in their accuracy and reduction in hallucinations, but it's confusing you because you think it's much more than it really is.
Also I'll say that when dealing with that 'obesity rate', there is value in absolute size, because per-capita numbers and things like that can be very badly skewed by extremely tiny populations.
Example: the YOY murder rate on Tonga is around 1/100k. But in 2018 there was a ridiculously silly spike:
https://www.macrotrends.net/global-metrics/countries/ton/tonga/crime-rate-statistics
In 2018 there was a double-murder: https://www.samoanews.com/regional/teen-sentenced-life-murder-couple-tonga
Now while that doesn't account for the full 9/100k, that means that that one singular act would've alone doubled the murder rate for the entire island (the population of Tonga is basically 100k so that makes this easy).
So per-capita is misleading in very tiny populations. Does that apply to obesity? Not really. But in general it actually does sometimes need context.
This is why you should >>>>NEVER<<<< trust anyone who only gives you information via percentages or words like 'double'. People who don't have data to back up percentages will just use percentages. People who don't have percentages to back up data will just give you data. If you see that, you're probably being manipulated.
Also I'm going to point out that if you ever find yourself 'arguing' with an AI, you're falling in to the mind trap. AI doesn't care. It doesn't listen. Your arguments are not recorded and processed and will change nothing about future interactions. Do not be deceived by these enigma engines.