Post

"I burned down a forest to give you the answer to the request you seek: Try mixing some glue in your pizza cheese!" - Google

📌

Sheeet. Maybe humans with experience in a topic and intelligence would be better? I know, shocking 😳

The easiest form of singularity, because it doesn't even have to reach 6th grade intelligence. :p

Not unique sure, but incorrect citations muddy the waters even further It makes the information look more reliable than it is. Instead of making search easier it makes it harder as you have to click on the citations and see if they actually say what the LLM summary claims

Seems grok doesn't grok very well.

Except the car gets worse and worse as technology changes and your car ages. I don't see how that relates well to what I posted. AI as a product is always innovating, and quickly. Imagine selling people on the IDEA of cars the same way, now that makes sense. Such a small post limit. Ridiculous.

Does it have use cases? Yes. But those cases are a very small subset of what it is being used for. It will **never** be able to do half the things they're trying to force it into right now. The sooner we stop trying to make this specialized tool do everything, the better.

Is that you, Sam???

I have that frustration daily. I feel like the exuberantly overconfident incorrectness is the most bothersome. At least show some uncertainty to clue me in that you might be off the raills.

it can be, but only if people are willing to improve it. discovery and progress cannot be erased - but it can be stalled and obselesced if we can't collectively validate its utility. i do foresee a future collabing w ML, but only if the intent of its purveyors align with us and the planet itself

Can we still call it artificial "intelligence" if its as thick as shit?

So they are accurately imitating actual human responses

By "the original article" I mean the article in Columbia Journalism Review, www.cjr.org/tow_center/w...

Then their creators shouldn't be putting them out there masquerading as search tools.

You mean they just make shit up instead of admitting they don’t know? No kidding.

I have caught incorrect information produced by Gemeni, and I likely would have only caught it because I already knew what I was asking it. They tend to correct if you point it out, but it's gonna be rough for people looking for information but aren't treading lightly.

Tough week for comrade Musk. His rocket blew up, Tesla's stock is falling faster than planes out of the sky under Secretary Duffy, and now to hear that Grok is dumb, just like him and the Toddler in Chief. He needs to jump around some more to get his energy back.

Glad you noticed before following the advice!

📌

Well luckily one of the largest industries on the planet hasn't invested hundreds of billions of dollars and tied its future to this technology. I will not be reading the replies.

No, I just mean that one of your own paragraphs is showing as if it's quoted (ie, in line with the LLM's paragraphs): it still sounds like you But yes, we do all start sounding like whoever or whatever we're reading!

Use Calude AI from Anthropic, its way way more reliable. Its the only one I trust the most for my kids & elderly patents apps.apple.com/us/app/claud...

I knew the template, just didn't know whose text it was. A good one!

A 60% failure rate? Seems like using those programs at this time would be foolish, if what you want is useful, data-driven analysis.

📌