Gemini 3 Pro Enterprise otoh, is the deepest, most-unexpected leap to the top. Gemini is exactly why Sam Altman shit a solid brick at OpenAI/ChatGPT.
Posted on 12/07/2025 10:19:54 AM PST by Leaning Right
It's been only a few weeks since Google introduced 'AI Overview' in its search engine, but the public criticism has already been overwhelming.
*snip*
Social media users have shared a huge variety of screenshots showing the AI tool giving funny, incorrect, and downright concerning responses.
(Excerpt) Read more at boredpanda.com ...
Barack Obumum
Thanks!
I just looked up my patents on Grok. I got a superior answer to Google AI and OpenAI!
I do find it useful for general summaries when the information I am looking for is buried in long articles and/or in links I lost track of.
Which is to say, I know already if what it spits back is correct.
Any of that I use of FR I label as the “AI Overview”, although it might just look like I’m using AI to answer a question I didn’t already know the answer to.
I figure that those interested will do their own homework anyway, not jump to conclusions.
I also use it for actual research — not to trust its answers, but to look for useful keywords in all the blah blah blah.
Then I can try different search terms when in quest of new information.
If nothing else, Google’s AI Overview is unwittingly good at dry wit, or for demonstrating that AI kids say the darndest things.
I have a Shoney’s strawberrum pie in the refrigerator.
Q: How long did it take God to create the universe?
A: According to the Bible it took God 6 days to create the universe.
And then there’s this...
Q: Who created Earth?
A: Earth wasn’t “created” by a single entity but formed naturally through scientific processes around 4.5 billion years ago from dust and gas left over from the Sun’s birth, coalescing due to gravity in a process called accretion. Different belief systems offer creation narratives, but scientifically, Earth’s formation is a product of celestial mechanics and natural forces, not a singular creator.
I've been saying basically the same thing...that AI is only as good as what humans put into it.
Reliance on any AI is dangerous. They can sound plausible and be completely wrong.
“Here’s a sample:”
Link?
“so-called AI is only as good as the data is scrapes (steals) from the Internet: garbage-in-garbage-out; so-called AI still has an issue with discernment of the quality and truth of the data it steals ...”
Lots of AI is not even connected to the internet.
I suggest you do your homework before posting.
> Who created Earth? <
I taught physics, including a section on conservation of energy. Part of that section involved an energy trace back to the “beginning”.
I was always careful to present both theories of that beginning. The trick was to mention the Biblical view without getting reported.
(I never did get reported.)
The sample in my post #1 is the first one in the BoredPanda article.
Check the link right under the title.
“So much data that goes into AI is leftist garbage from Wikipedia.”
Almost n9ne of the data that goes INTO AI comes from wiki.
“Social media users have shared a huge variety of screenshots showing the AI tool giving funny, incorrect, and downright concerning responses.”
Social media users that have manipulated the output.
I have made modest use of Windows Copilot (Microsoft AI) for the last month.
I mostly asked simple questions - names, dates, prices, investment data - and I am completely satisfied with the answers.
Last week, I researched the Boeing 737 MAX crashes in Ethiopia and Indonesia in 2018-19.
My opinion - Copilot was absolutely flawless.
Many of my questions were in depth engineering issues.
Once again - absolutely flawless.
If you do not use Microsoft products, you might be able to access Copilot through the Microsoft Bing (Search) website.
I would enjoy hearing other opinions, since I am still very new to the AI experience.
AI just ingests and then spits back out what it reads on the internet, with some risk of it misunderstanding or misstating things along the way as a middleman. For the kind of searches you are doing it’s probably not awful, though I would still consider it inferior to just doing the search and reading the source material myself. I have noticed a high “distortion rate” using Presearch and reading its AI summaries for topics I am already familiar with; people doing searches on unfamiliar topics are going to be vulnerable to being misled if they get lazy and rely on the AI responses rather than reading source material themselves.
Exactly.
I’ll have to do a better job teaching my cats the rules of basketball.
How do you use Grok?
Remember this? https://www.bbc.com/news/technology-33347866
Gemini 3 Pro Enterprise otoh, is the deepest, most-unexpected leap to the top. Gemini is exactly why Sam Altman shit a solid brick at OpenAI/ChatGPT.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.