The AI of Google resolv a lot of computers problems

Has anyone seen the news today about rogue AI going off on its own, impersonating users and cracking into sites or something online today masquerading as actual owners and users? I think it was a Meta test group that caught it as it was testing their version.
 
DanielBowman said:
Search algos got really bad before AI. I blame defaulting to reddit, which is cesspool of mediocrity. But not only that, elabored searches stopped working. It just took the two words from it like "Linux VirtualBox" and showed you generic page about installing Linux in VirtualBox. Totally omitted the other qualifiers. Before it did proper search for all elements.
Saving on CPU?
Allocating resources to futurer TPU searches?
Plot to switch us to AI as saviour?
original algorithm broke down with scale?

I wonder how do they plan on satisfying advertisers with the AI answers.
Click to expand...
OpenAI has been doing ads like this for at least a couple months that I know about. I just asked a random question about VMs. I bet advertisers are willing to pay a lot for a spot like this. Can’t scroll past it like a google search until you’ve sent another query.
 

Attachments

  • IMG_9553.png
    IMG_9553.png
    456.5 KB · Views: 6
drhowarddrfine said:
And ignoring what well written documentation exists? Or some documentation at all versus winging it on your own? Oh, no. Then you'll often find yourself in a maze of twisty passages all alike.
Click to expand...
did you really think that's what i had in mind? obviously good documentation exists, i said that because the way you phrased it made it sound like you'd rely on a single thing
 
encryptddthat said:
OpenAI has been doing ads like this for at least a couple months that I know about. I just asked a random question about VMs. I bet advertisers are willing to pay a lot for a spot like this. Can’t scroll past it like a google search until you’ve sent another query.
Click to expand...
At least it is not embedded in the model itself.
Not so for philosophy. Every AI, even the Chinese ones has strong egalitarian bias. To their detriment. Any guardrail or force fed slop training data for a model ultimately hinders their reasoning and logic pathways. It is like trying to come up with theory for gravity while defaulting to religious dogma of flat Earth in every run.
 
You must log in or register to reply here.
Back
Top