• Welcome to the Internet Infidels Discussion Board.

Immigration Issues

ICE raids another car wash—target U.S. citizens.

"I got my ID! Check my locker!" worker yelled as agents twisted him to ground.

"I thought it was active shooter and hid in bathroom with 5 other guys," said customer.

"Everybody started screaming. My car is still sitting with towels on it," a man said.

The manager said they just grabbed anyone they could—without asking for any identification.

Agents even took customers' keys—leaving them stranded.

The incident occurred at the Studio City Hand Car Wash in Los Angeles, California.


A bit more detail and less misleading. It's still horrible.


 
Sorry, but please, NO! Don't hunt them with AI, it will finger innocents. AI will give you shit rather than say "I don't know."
Really? I haven't had this experienced yet. It has for example on occasion required me to discern whether or not a Reddit post was a good source of information or in general to question it's links. Something I would do anyways. So yeah, some action on the user's part still required. Don't expect it to be all things so early in it's young life. Let it learn to walk first.
Never seen it???

Simple one I hit and remembered:

"largest tree on mt charleston nevada"

Google's AI response:
"The "largest" tree on Mt. Charleston is often considered the Rain Tree, a 3,000-year-old Great Basin Bristlecone Pine with a massive 12-foot circumference and 455-inch diameter, making it the most massive bristlecone pine in the state. While the exact location of the single largest bristlecone pine in Nevada is a closely guarded secret, the Rain Tree is a famous and accessible example of these ancient, monumental trees in the region."

Self-contradictory, so obviously false. I've actually been at the Raintree many times, it's not 12' in diameter. I won't blame Google for that 3,000 year old bit, that's the common knowledge of those of us who frequent the area but it turns out to be false, the supposed root source said they had no cores of it and were not aware of any trees in the area having been cored. And finding the coordinates of the actual 12' tree wasn't very hard--but getting to it doesn't look easy.

When reading this article, I assumed these ICE agents were being identified in a very similar way masked individuals on Jan 6th were identified, that the technology is just more in the public sphere and probably much more accurate all these years later.

Once a name is associated with a masked face, the rest of an individual's info is easily gathered from county websites and data brokers with consumer facing websites. So whether or not an individual is a match is no more difficult than the commonality of the person's name.
I'm fine with outing ICE agents, I just don't trust AI to be accurate about it.
 
Sorry, but please, NO! Don't hunt them with AI, it will finger innocents. AI will give you shit rather than say "I don't know."
Really? I haven't had this experienced yet. It has for example on occasion required me to discern whether or not a Reddit post was a good source of information or in general to question it's links. Something I would do anyways. So yeah, some action on the user's part still required. Don't expect it to be all things so early in it's young life. Let it learn to walk first.
Never seen it???

Simple one I hit and remembered:

"largest tree on mt charleston nevada"

Google's AI response:
"The "largest" tree on Mt. Charleston is often considered the Rain Tree, a 3,000-year-old Great Basin Bristlecone Pine with a massive 12-foot circumference and 455-inch diameter, making it the most massive bristlecone pine in the state. While the exact location of the single largest bristlecone pine in Nevada is a closely guarded secret, the Rain Tree is a famous and accessible example of these ancient, monumental trees in the region."

Self-contradictory, so obviously false. I've actually been at the Raintree many times, it's not 12' in diameter. I won't blame Google for that 3,000 year old bit, that's the common knowledge of those of us who frequent the area but it turns out to be false, the supposed root source said they had no cores of it and were not aware of any trees in the area having been cored. And finding the coordinates of the actual 12' tree wasn't very hard--but getting to it doesn't look easy.
B – See, in my mind these two things are not equal. There is a chasm of context between what I wrote and your interpretation of it. Sure, if one strips it down to its fundamental meaning, one could argue these two statements are equal. But that's not how the English language is suppose to work. It's not some problem to be solved by tossing all statements in one of two boxes. There implication and distinction, gradations we must interpret.
 
My google ai says nothing about the tree's diameter or circumference. And it always includes a link to where it obtained the information.
 
Last edited:
ICE detains green card holder—days after she passed citizenship exam.

She was awaiting the official ceremony—when agents tricked her into coming in for appointment—only to arrest her.

Sharareh Moghadam entered U.S. legally with all documents in order—but recent trip to Iran may have flagged her in ICE database.

She and her husband have owned their Los Angeles balloon shop together for almost 20 years.

Hooshang Aghdassi hasn't seen his wife in over a month—after she was transferred to an ICE detention center in Phoenix, Arizona

 
There is a chasm of context between what I wrote and your interpretation of it. Sure, if one strips it down to its fundamental meaning, one could argue these two statements are equal. But that's not how the English language is suppose to work. It's not some problem to be solved by tossing all statements in one of two boxes. There implication and distinction, gradations we must interpret.
That's what all astrologers and psychics say when challenged.

I am not wrong; You just don't understand the subtleties of my craft.

AI is the new charlatan on the block. It's highly convincing, and it's even right more often than it is wrong. That doesn't make it reliable. Or valuable.
 
Sorry, but please, NO! Don't hunt them with AI, it will finger innocents. AI will give you shit rather than say "I don't know."
Really? I haven't had this experienced yet. It has for example on occasion required me to discern whether or not a Reddit post was a good source of information or in general to question it's links. Something I would do anyways. So yeah, some action on the user's part still required. Don't expect it to be all things so early in it's young life. Let it learn to walk first.
Never seen it???

Simple one I hit and remembered:

"largest tree on mt charleston nevada"

Google's AI response:
"The "largest" tree on Mt. Charleston is often considered the Rain Tree, a 3,000-year-old Great Basin Bristlecone Pine with a massive 12-foot circumference and 455-inch diameter, making it the most massive bristlecone pine in the state. While the exact location of the single largest bristlecone pine in Nevada is a closely guarded secret, the Rain Tree is a famous and accessible example of these ancient, monumental trees in the region."

Self-contradictory, so obviously false. I've actually been at the Raintree many times, it's not 12' in diameter. I won't blame Google for that 3,000 year old bit, that's the common knowledge of those of us who frequent the area but it turns out to be false, the supposed root source said they had no cores of it and were not aware of any trees in the area having been cored. And finding the coordinates of the actual 12' tree wasn't very hard--but getting to it doesn't look easy.
B – See, in my mind these two things are not equal. There is a chasm of context between what I wrote and your interpretation of it. Sure, if one strips it down to its fundamental meaning, one could argue these two statements are equal. But that's not how the English language is suppose to work. It's not some problem to be solved by tossing all statements in one of two boxes. There implication and distinction, gradations we must interpret.
Actually, AI needs to interpret... and it isn't there yet. Of course, AI to compare images is a bit different than the Google AI out there. AI as it exists now is a tool... that should be used as one of many tools especially when it comes to identifying people.

Now if you'll excuse me, I need to work on getting that Nobel Prize for redefining the value of Pi based on the Google AI search above.
 
There is a chasm of context between what I wrote and your interpretation of it. Sure, if one strips it down to its fundamental meaning, one could argue these two statements are equal. But that's not how the English language is suppose to work. It's not some problem to be solved by tossing all statements in one of two boxes. There implication and distinction, gradations we must interpret.
That's what all astrologers and psychics say when challenged.

I am not wrong; You just don't understand the subtleties of my craft.

AI is the new charlatan on the block. It's highly convincing, and it's even right more often than it is wrong. That doesn't make it reliable. Or valuable.
It's explanation and links (ChatGPT) provide a very good jumping off point. That has value to me and likely to many others. Maybe it will get better. Maybe it will just bounce sideways for the next forty years like a Microsoft operating system, forever finding new and interesting ways to suck. We don't know. But to dismiss it at this point as a charlatan is premature.
 
There is a chasm of context between what I wrote and your interpretation of it. Sure, if one strips it down to its fundamental meaning, one could argue these two statements are equal. But that's not how the English language is suppose to work. It's not some problem to be solved by tossing all statements in one of two boxes. There implication and distinction, gradations we must interpret.
That's what all astrologers and psychics say when challenged.

I am not wrong; You just don't understand the subtleties of my craft.

AI is the new charlatan on the block. It's highly convincing, and it's even right more often than it is wrong. That doesn't make it reliable. Or valuable.
It's explanation and links (ChatGPT) provide a very good jumping off point. That has value to me and likely to many others. Maybe it will get better. Maybe it will just bounce sideways for the next forty years like a Microsoft operating system, forever finding new and interesting ways to suck. We don't know. But to dismiss it at this point as a charlatan is premature.
On the contrary, it is inherent and unavoidable in the way LLMs work.

They present information from a wide variety of un-vetted sources in a manner and style that carries all the hallmarks of knowledge amd trustworthiness. But they are neither knowlegable nor trustworthy.

They are, by design, highly skilled fraudsters and charlatans.
 
My google ai says nothing about the tree's diameter or circumference. And it always includes a link to where it obtained the information.
Interesting. Now I'm getting a very different and more accurate response to the exact same query I did before.
 
On the contrary, it is inherent and unavoidable in the way LLMs work.

They present information from a wide variety of un-vetted sources in a manner and style that carries all the hallmarks of knowledge amd trustworthiness. But they are neither knowlegable nor trustworthy.

They are, by design, highly skilled fraudsters and charlatans.
Yeah. Especially with that 3,000 year bit--common knowledge that is wrong.
 
:hysterical: ICE was gathering in a federal building parking lot preparing to go out for raids and someone used a bike chain on the gate and locked them in.

 
  • Like
Reactions: WAB
Back
Top Bottom