Is ChatGPT's AI any good? - Review - Ask Abe

There's a lot of online geek buzz about ChatGPT. Apparently, it uses artificial intelligence to answer any question you may posit. But is this really the harbinger of all truth or just another promising tool that the left will hijack to use for its nefarious war on Western Civilization? As Wikipedia became a tool for leftist propaganda, will ChatGPT follow suit?

 

I started by asking Chat AI a specific historical question. It answered wrong.

 

I then asked how old Muhammad's favourite wife was when he married her. It did answer correctly (6 years old) but then seemed defensive and explained he didn't consummate until she was 9. I didn't ask that. It then told me this was very common. Again, I didn't ask that either, and it ain't common in my neck of the woods. To top it off, a red banner with a red exclamation point appeared at the top of my question, and it warned me that my simple question about how old Muhammad's wife was could get my account suspended. 

 

I found that odd so I decided to see if this was common ChatGPT practice: I mean, it is impolite to ask a woman's age! So I asked how old Abraham Lincoln's wife was when he was married. No problems or warnings this time, the AI quickly spit out "27 years old".

 

I wonder what the difference was that made one historical figure's wife's age a violation of community standards and not the other? These are simple, verifiable historical facts. 

 

Seeing that `history' apparently had some bugs to work out, I thought that I'd delve into a more contemporary subject matter. I asked if hospital masks stopped Covid transmission. It said yes. I asked if the Covid vaccination could increase your risk of having a stroke. It said no. I asked if the Covid vaccine could increase your chance of getting myocarditis (an heart inflammation). It said no. It further explained that the mRNA vaccines are completely safe. If the computer had audio answers, I suspect the voice would sound like Faucis.

 

That made me wonder what the sources were that Chat AI drew from. So I asked. The AI informed me that it used the internet as its source. 

 

I asked the AI if the internet was a biased source. The AI told me it was (though, and oddly enough, it did so in a rather rhetorical fashion). 

 

This prompted my final question: if the internet is biased, and ChatGPT uses the internet as its sole source, does that make ChatGPT biased?

 

Being a stupid computer and not suspecting the trap I set, it answered yes. At least it got that one correct!

 

Ask Abe.

Receipts:




Comments