ChatGPT 3.5 vs ChatGPT 4.0 vs Bard vs Fact

Like many others I have been using ChatGPT 3.5 (plus). I have also tried 4.0 and now Bard as well.

Having tried them all at least a little, I was curious on the differences in answers I would get.

So, I decided to conduct a little experiment.

I used a question that was differes from year to year and that is 100% verifiable.

I entered the exact same question in the exact same words on all three to see what would happen.

1. Would I get the correct answer?

2. Would they provide additional information that might be helpful?

Question: what is tax day 2023

Verifiable Answer: The IRS states... "April 18 tax filing deadline in 2023"

ChatGPT 3.5

1. Did ChatGPT 3.5 get the correct answer? Sadly, no.

2. Would they provide additional information that might be helpful? It did provide additional informatiopn that is justifying the reasoning behind the change in date but it mistakenly placed the Emancipation Day holiday on the 16th instead of the 17th. This in turn made the information incorrect.

ChatGPT 4.0

1. Same basic outcomes a ChatGPT 3.5.

2. Same basic outcomes a ChatGPT 3.5.

Bard

1. Would Bard get the correct answer? Yes. Bard did get the correct answer and made the correct correlation for the holiday.

2. Would Bard provide additional information that might be helpful? The additional information provided is interesting and informative.

IRS

I hope you found my little experiment interesting, as I did.

My Take Away: I would suggest to always double check the accuracy of the information that you get.

What do you think?

Please let me know in the comments below. (ha ha, I can rhyme)