artificial intelligence

Asproule1984 · Post by **Asproule1984** » August 31st, 2023, 7:28 am

I have been hearing rumblings that Librivox has been selling voices to be used to train AI. If this is true, I can only speak for myself when i say that I do not consent to this unless I am getting paid royalties. I am 1000% sure other volunteers would agree.

Post by **TriciaG** » August 31st, 2023, 7:46 am

Absolutely false. We don't sell anything and never will.

However, these recordings have been released into the public domain, so AI trainers can and will use them without your permission, for free.

Asproule1984 · Post by **Asproule1984** » August 31st, 2023, 8:33 am

Right. Was not trying to be reactionary. Can't remember the specifics, but heard rumblings about companies messing with internet archive. That and some of the ethically sketchy stuff surrounding AI had me concerned. I really enjoy volunteering here

GettingTooOld · Post by **GettingTooOld** » August 31st, 2023, 9:10 am

Asproule1984 wrote: ↑August 31st, 2023, 8:33 am Right. Was not trying to be reactionary. Can't remember the specifics, but heard rumblings about companies messing with internet archive. That and some of the ethically sketchy stuff surrounding AI had me concerned. I really enjoy volunteering here

Well, not trying to be alarmist but AI will kill us all

AI grabs everything on the internet that you can reach and a great deal that you don't know is on the internet. That's in addition to a huge amount of
"private" data. Like medical data.

You also contribute to it in ways you do know about, like answering capchas which is used to train AI. Looking at history and future, unfortunately there are no 'three laws' for robotics, drones have never had that and never will. It's confined to fiction and wishful thinking. Drones do what drones do, and they're told to do things that are not nice. Now, they'll know everything that you click on, so they'll know which pictures contain a schoolbus, so there will be a checkbox for target/don't target school buses and so on.

Ultimately AI does only what people design it to do, same as any tool. People make nasty evil tools and can use them because the rest of us don't object strongly enough. Evil triumphs when good people do nothing. Always has been the way, always will be the way.

There is no way to stop AI being trained on data, as the developers doing it won't be publishing what is going on. The only limitations possible are against using your work in new work is where you can see some of the original, and is the derivative is subsequently published.

To stop that, you'd use a creative commons restricted license, which will let people use your work for non-commercial purposes if you choose the right one of the licenses.

For public domain, there's no limit as to how it can be used. What you may do, which is constructive, is to realize it's more the inevitable bad uses of AI that are the problem, and that is the thing that must be stopped, and the evil people who will use it against us all. If people make no plans to limit evil, it will triumph.

People power is in short supply and it's bringing the world to the brink of destruction. Alarmist ? meh, maybe, I mean when you're in a building that's on fire and about to be flattened by lavaflow if the pile of C4 explosive doesn't go off first and the Tsunami doesn't put out the flames, then maybe alarmist is not a 'bad' thing.

Or we can be hysterically anti-alarmist. There's always that.

Post by **TriciaG** » August 31st, 2023, 9:14 am

And with that, we'll close this thread. The OP's question is answered, and we aren't going to start a debate about AI and its possible future.