• Welcome to the Internet Infidels Discussion Board.

“Generative AI is a sack of wet garbage”

 
 
 
My suggestion: get off the internet, and especially Google, and read a book or two sometime by people who have actually studied their fields and know what they are talking about. Such a quaint idea.
 
In another thread I noted how AI confused a college friend, Ira Elliott, a Hemingway scholar (note the spelling of the last name) with Ira Elliot (note the different spelling of the last name) the Nada Surf drummer. AI apparently believed that the Hemingway scholar moonlighted as a drummer. How much lower can we sink until we vanish? :unsure:
 
I’ve been noting this relatively new word for AI, that is “scrapes” stuff off the internet. I guess this is a polite euphemism for “steals.”

So it steals scrapes stuff off the internet, much of it copyrighted and to which its makers have no legal right, and then recombines it into a bunch of BS that Google shoves down our throats.

Brave New World!
 
In fact, I believe almost EVERYTHING on the internet is automatically copyrighted, including posts on this board.

The only things not copyrighted are writings that are of a certain longer age.
 
pood said:
In fact, I believe almost EVERYTHING on the internet is automatically copyrighted, including posts on this board.

The only things not copyrighted are writings that are of a certain longer age.
Click to expand...
The problem with this, with respects to AI, is that the copyright of this board does not preclude readers from reading the content here. Anyone can wander in, read what content is publicly available, have some experience of having read it as many times as their awareness skates across the whole of it for however long a browser will retain the text of it, and then to be done with it.

An AI does the digital equivalent of that when it does training.


However, some acts of clear piracy have been done by AI creators: book piracy and video piracy and the access of content through BitTorrent and so on.

That has been done on a massive scale, but then... I have pirated more books than I have paid for, myself.


Also, I hate to mention this, but this also doesn't prevent someone from doing all this within the bounds of copyright.

The sum total of information that is absolutely needed to train an AI to a reasonable human degree doesn't really exceed the amount of information needed to do this with a human, or at least I don't see why it would.

I have had what, maybe a few hundred thousand dollars worth in books and education? At most 500k?

A lot of my time has instead been spent playing video games, or talking here... Which as noted anyone can read the records of.


Which gets me to my point: very little of this is strictly a copyright violation in the first place.


So in an interest of not wanting to pass a law outlawing "robots" from reading or whatever, I think we shouldn't consider reading the open internet to be a crime, especially when the system has a subscription, library card, or account.


I do also think, however, that letting an AI read literally everything and anything and expecting it to be capable of "predicting" the material is a fool's errand because most books are trash, most websites are trash, most forum posts are trash, most news is trash.

Pushing an unmitigated pile of trash that deep onto any sort of entity will result in something that is very good at creating and understanding... Trash.

So of course most of it is just... Trash. Because you are what you eat. And we fed this thing mostly trash.
 
unapologetic said:
AI on YouTube is out of control:
This NEEDS to be marked as fiction
Click to expand...
I watched a YouTube video featuring George Will. I've been reading George's columns and watching him on TV news for many years and something looked and sounded a little off. The first thing was the audio speed. He was talking pretty fast and his head didn't look like it fit his body. What tipped me off was he was talking about things that happened a month ago as if it was today. The video had been posted 20 minutes earlier. I went to the channel's home page, where they admitted it was an AI creation. George probably did say all those things, not in the same order, and not as lately as implied. There are at least three George Will channels that have all the authenticity of an Elvis tribute band. Same goes for George Conway. It must be a George thing.
 
I also saw a George Will vid. Some clickbait title like 'He signed the 'Removal order'.
I wanted to know what a removal order was, who signed, who was removed, from what. (hopeing it was removing Rump)
It didn't answer any of those questions. Just went on and on about the importance of government fallowing correct procedures, and being very repetive. Obviously written by AI. I've also seen an AI Jimmy Kimmel. Is this leagle? Can't they sue?
 
When Apollo 11 landed on the moon, there was a speech prepared for President Nixon just in case the mission failed and the astronauts died on the moon. There is a deepfake video of Nixon reading the speech (never given, obviously) from the Oval Office. It’s indistinguishable from something real, but in this case we know it couldn’t have been real. That will not always be the case. In a few years we may be unable to distinguish reality from fiction. The sci-fi writer Philip Dick anticipated all this decades ago.
 
bilby said:
pood said:
In a few years we may be unable to distinguish reality from fiction.
Click to expand...
For many centuries, many people have been unable to distinguish reality from fiction. This is just more of the same.
Click to expand...

True enough. :ROFLMAO: We see people on this very forum unable to do that. Shit like AI and deepfakes will only worsen the problem, though.
 
The Singularity Is Near: When Humans Transcend Biology is a 2005 book by futurist Ray Kurzweil that argues technology will soon merge with humans, leading to a "Singularity" where our intelligence becomes nonbiological and vastly more powerful, allowing us to overcome biological limitations like aging and death. Based on the "law of accelerating returns," Kurzweil predicts this era of radical transformation, driven by exponential growth in information technology, will occur around 2045
Click to expand...

If I submit "Kurzweil singularity" to Google's n-gram viewer, I see it first appeared in 2009 but plateaued after 2018. Has Kurzweil's prediction become obsolete with the Singularity arriving long before 2045? (And much less benign than Kurzweil imagined.)

unapologetic said:
AI on YouTube is out of control:
This NEEDS to be marked as fiction
CQ67mrnmE_Q
Click to expand...

YouTube has "rules" and there is an "Altered or synthetic content" that MIGHT appear near the upper left when you start the video. (I write "MIGHT" because its presence seemed erratic and I didn't try to deduce the algorithm for its presentation.)
 
pood said:
When Apollo 11 landed on the moon, there was a speech prepared for President Nixon just in case the mission failed and the astronauts died on the moon. There is a deepfake video of Nixon reading the speech (never given, obviously) from the Oval Office. It’s indistinguishable from something real, but in this case we know it couldn’t have been real. That will not always be the case. In a few years we may be unable to distinguish reality from fiction. The sci-fi writer Philip Dick anticipated all this decades ago.
Click to expand...
1767153667114.png
 
bigfield said:
pood said:
When Apollo 11 landed on the moon, there was a speech prepared for President Nixon just in case the mission failed and the astronauts died on the moon. There is a deepfake video of Nixon reading the speech (never given, obviously) from the Oval Office. It’s indistinguishable from something real, but in this case we know it couldn’t have been real. That will not always be the case. In a few years we may be unable to distinguish reality from fiction. The sci-fi writer Philip Dick anticipated all this decades ago.
Click to expand...
View attachment 53205
Click to expand...
Well, there's the technology out there capable of resolving what images are real and what images are not, but the trick is, it has to be verified/verifiable at the point of image creation, and some technologies can be contrived to argue for letting perfect be enemy of the good.

Analog film cameras might make a comeback, even.

People will figure out how to record the present and preserve the past regardless of what fakery will come in the near term.
 
Jarhyn said:
bigfield said:
pood said:
When Apollo 11 landed on the moon, there was a speech prepared for President Nixon just in case the mission failed and the astronauts died on the moon. There is a deepfake video of Nixon reading the speech (never given, obviously) from the Oval Office. It’s indistinguishable from something real, but in this case we know it couldn’t have been real. That will not always be the case. In a few years we may be unable to distinguish reality from fiction. The sci-fi writer Philip Dick anticipated all this decades ago.
Click to expand...
View attachment 53205
Click to expand...
Well, there's the technology out there capable of resolving what images are real and what images are not, but the trick is, it has to be verified/verifiable at the point of image creation, and some technologies can be contrived to argue for letting perfect be enemy of the good.

Analog film cameras might make a comeback, even.

People will figure out how to record the present and preserve the past regardless of what fakery will come in the near term.
Click to expand...
Polaroid camera film is still in production. The primary users are people who document crime scenes. Unlike film and digital camera photos, Polaroid snapshots are still very difficult to manipulate, so they are accepted as a true depiction of whatever the camera was pointed at.
 
Bronzeage said:
Jarhyn said:
bigfield said:
pood said:
When Apollo 11 landed on the moon, there was a speech prepared for President Nixon just in case the mission failed and the astronauts died on the moon. There is a deepfake video of Nixon reading the speech (never given, obviously) from the Oval Office. It’s indistinguishable from something real, but in this case we know it couldn’t have been real. That will not always be the case. In a few years we may be unable to distinguish reality from fiction. The sci-fi writer Philip Dick anticipated all this decades ago.
Click to expand...
View attachment 53205
Click to expand...
Well, there's the technology out there capable of resolving what images are real and what images are not, but the trick is, it has to be verified/verifiable at the point of image creation, and some technologies can be contrived to argue for letting perfect be enemy of the good.

Analog film cameras might make a comeback, even.

People will figure out how to record the present and preserve the past regardless of what fakery will come in the near term.
Click to expand...
Polaroid camera film is still in production. The primary users are people who document crime scenes. Unlike film and digital camera photos, Polaroid snapshots are still very difficult to manipulate, so they are accepted as a true depiction of whatever the camera was pointed at.
Click to expand...
I know some folks who rock Polaroid cameras.

They're all HUGE nerds for Polaroid.
 
bilby said:
pood said:
In a few years we may be unable to distinguish reality from fiction.
Click to expand...
For many centuries, many people have been unable to distinguish reality from fiction. This is just more of the same.
Click to expand...
Most of the fakes out there aren't high quality--they don't need to be to fool the average person inclined to support the position.
 
You must log in or register to reply here.
Back
Top Bottom