I was an early beta tester of Google's Bard AI

I was an early beta tester of Google's Bard AI, now known as Gemini. Back then I was excited about a chatbot that could provide concise results to my search engine questions. It was soon apparent that the majority of the answers Bard returned were incorrect by varying degrees. Fast forward a number of years and Gemini results aren't much better. For example, I asked it for the standard tuning for a Puerto Rican Cuatro stringed instrument and the original result gave me the tuning for six strings even though the instrument has ten strings. I had to correct multiple times before it finally acknowledged my answer.

The amount of time and energy being wasted on this trash is infuriating.

53 points · 4 comments · view on lemmy.world

4 Comments

sunzu2@thebrainbin.org · 16 pts · 1y (1 reply)
[ removed ]
JoMiran@lemmy.ml · 9 pts · 1y

💯

ricecake@sh.itjust.works · 7 pts · 1y (2 replies)

So stop using it? If you keep using it they're just going to think you want it, doubly so if they're not getting the feedback that the answer was trash.

JoMiran@lemmy.ml · 13 pts · 1y (1 reply)

It is built into my phone. Even though I use Firefox as a browser and Kagi for search, sometimes Google finds a way to weasel Gemini in there and have it give me results. That's what happened today, and the results were so wrong that I had to stop what I was doing and do a fresh round of tests. Just as it was years ago, the results are mainly trash.

whats_all_this_then@lemmy.world · 4 pts · 1y

I miss the good old days when software shoving itself where it didn't belong against your will immediately got it classified as malware