Bard the Google AI is just as silly as the other AI programs

AI Thread Summary
The discussion highlights the limitations of Google's AI, Bard, in accurately retrieving and presenting academic research information. When asked to list research papers by a PhD supervisor, Bard provided a very limited selection, failing to include any papers published after 1975 and incorrectly stating the supervisor's status. After clarification, Bard acknowledged the supervisor was alive and corrected its information. Additionally, when prompted about the user's own publications, Bard listed incorrect titles and details. The conversation emphasizes that while AI can produce grammatically correct responses, it struggles with context and factual accuracy, suggesting that users need to formulate precise questions to receive better answers. Overall, the experience illustrates the ongoing challenges in AI's ability to handle basic factual inquiries effectively.
DrJohn
Messages
212
Reaction score
192
TL;DR Summary
Reported someone as dead, claimed he died in 1975, but was still publishing in 1980
So I got an invite to try Bard, Google's AI machine.
I asked it to list research papers by my PhD supervisor in the area we had worked on. It listed just five (clearly a very small subset) with none published after 1975.

So I asked again but this time asked for papers after 1975. Again it listed only five papers, a small subset that went up to 1980.

Then it said that he had passed away in 1975 and that the papers it had just listed had been published before he died, in 1975. Strange given that there was one for each year of 1976 - 1980.

When I pointed out my old supervisor was still alive, it apologised and told me he was alive, and listed correctly his current status and which Uni he was working at (I checked and it was correct this time).

Looks like Google is having the same problems with lying about the living and the dead. Anyone else tried Bard yet? Any better results than mine?
 
Computer science news on Phys.org
For a laugh, just in case I had died too and not noticed, I asked it if I had published any papers in the field I had researched. It replied listing only three papers, all with the wrong title ! and one in the wrong journal but with the correct page number!
 
  • Wow
Likes FactChecker
For better answers just as in this forum, one needs to make a better question and focus on assumptions and context. Yet for elementary common sense only a few words are needed. likewise, We can’t blame professors for wrong answers but they would not make such obvious faux pas, yet it may have perfect spelling and syntax.

It’s not perfect, nor am I.
 
In my discussions elsewhere, I've noticed a lot of disagreement regarding AI. A question that comes up is, "Is AI hype?" Unfortunately, when this question is asked, the one asking, as far as I can tell, may mean one of three things which can lead to lots of confusion. I'll list them out now for clarity. 1. Can AI do everything a human can do and how close are we to that? 2. Are corporations and governments using the promise of AI to gain more power for themselves? 3. Are AI and transhumans...
Thread 'ChatGPT Examples, Good and Bad'
I've been experimenting with ChatGPT. Some results are good, some very very bad. I think examples can help expose the properties of this AI. Maybe you can post some of your favorite examples and tell us what they reveal about the properties of this AI. (I had problems with copy/paste of text and formatting, so I'm posting my examples as screen shots. That is a promising start. :smile: But then I provided values V=1, R1=1, R2=2, R3=3 and asked for the value of I. At first, it said...

Similar threads

Replies
25
Views
4K
Replies
1
Views
2K
Replies
11
Views
3K
Replies
3
Views
563
Back
Top