- Joined
- Sep 6, 2019
- Messages
- 23,639
- Reaction score
- 121,675
- Location
- you tell me and we will both know
What we have right now isn't even AI. It's an LLM, which just predicts what word people want to see next, sort of like autocomplete. It has no conception of whether the data it is giving is accurate.AI is not a useful tool for this, as evidenced by references to Weaver maximizing Holland. I've recently asked similar questions in the context of trade partners, and it didn't seem to take into account the last 2 drafts, much less the latest transactions. After repeatedly feeding it data and reframing my questions, I gave up in frustration. It felt it should be paying me, and not the reverse. That's Chatgpt, not Grok, but the point stands.
My favourite part is when I chastize it in jest and it comes back with "Apologies, I can see why you would be frustrated. Here's a 100 % correct response". Which is always another abomination.What we have right now isn't even AI. It's an LLM, which just predicts what word people want to see next, sort of like autocomplete. It has no conception of whether the data it is giving is accurate.
Exactly. Their language model isn't trained on in depth NFL data. That's the very issue with AI. You get hallucinations if you are using an algorithm that isn't appropriate for your question(s). There will need to be numerous models if you want specific deep dive info on various subjects...very expensive!The problem with using "AI" for that type of thing is that algorithms, for one, do not process context very well to start with. With questions like this, there's really no factual database with which to draw on either. All it can really do is search web articles and arrive at an attempted consensus of those, usually biased, opinions.
I think most ppl have a mistaken idea about what AI is or how it works.
Don't get me wrong, for some things it is awesome and making great improvements at a rapid pace. It's amazing at massive amounts of data processing and response formulation.
There is no underlying actual "intelligence" by common definition, coming to abstract conclusions involved yet, though. It's just a matter of time.
Garbage in = garbage out
Dan Marino dead for 60 years, but cloned from a strand of DNA and brought back.I dont know when the date will be maybe 2059?
But these are the guys who will finally break the curse according to the Grok
View attachment 191880View attachment 191881
Spot on assessmentDan Marino dead for 60 years, but cloned from a strand of DNA and brought back.
Finally gets his SB, even though he looks like Derek Carr.
Even better when it's the same answer with like one word changed. I find it fairly unbelievable that people actually use LLMs for anything important.My favourite part is when I chastize it in jest and it comes back with "Apologies, I can see why you would be frustrated. Here's a 100 % correct response". Which is always another abomination.
Same!I didn't have my glasses on first time I read this thread title, I said "Why the hell would anybody ask Gronk about the Phins". Geez, I cannot wait until camp starts
Playoffs? Playoffs??Lol at you guys saying AI is in inconsistent etc.
I read the the thing. Th first parts are true and redundant and look like every article written about the team the last couple years. The conclusion is spot on.
They sr my going to any SB this year or next. Might get playoffs.
It's absolutely correct.