News for hackers. Stuff that matters.

A comment on Gemini: Google's most capable AI model yet. Parent comment

Re:Gemini: Google's most capable AI model yet by spaceman_2020 on Wednesday December 06, 2023 @03:22PM (#38548979)
No, I just use my airpods and the iPhone app. I use the voice chat and ask questions, then follow up questions.

Like I was reading a book about Genghis Khan the other day, which got me into Wikipediaing all his sons and grandsons. While doing chores, I asked chatGPT "tell me about Genghis Khan's sons", and then follow up questions like "what was the name of the khanate or empire Ogedai founded?"

It's an absolutely fantastic tool for diving into a new topic.

[ Parent | on HN ]
Re:Gemini: Google's most capable AI model yet by p_j_w on Wednesday December 06, 2023 @05:03PM (#38550151)
>While doing chores, I asked chatGPT "tell me about Genghis Khan's sons", and then follow up questions like "what was the name of the khanate or empire Ogedai founded?"

You actually trusted the information it gave you?

[ Parent | on HN ]
Re:Gemini: Google's most capable AI model yet by spaceman_2020 on Thursday December 07, 2023 @01:36AM (#38553526)
If I trust Wikipedia for this info, then I have no reason to distrust it.

And I’m asking about Ogedai Khan and the Ilkhanate, not my child’s health. It’s okay if I learn something wrong :D

[ Parent | on HN ]
Re:Gemini: Google's most capable AI model yet by kolinko on Thursday December 07, 2023 @04:23AM (#38554504)
With time you learn where it’s trustworthy and where you need to check for bullshit.
[ Parent | on HN ]
Re:Gemini: Google's most capable AI model yet by m_fayer on Wednesday December 06, 2023 @04:41PM (#38549908)
I think the Star Trek computer has actually been achieved now. And Google didn't get there first.
[ Parent | on HN ]