I dont mind AI voice if it wasnt an indicator that the script was written by chatgpt. Some people are good at all the behind the camera stuff and I dont necessarily think they should have to find a voice actor for a video essay. I dunno how much data centre madness is involved in glorified text to speech, any amount is too much there. But I get the accessibility option of text to speech and get not wanting to sound like Stephen Hawking while using it.
There’s a ton of really good programming channels that use AI voice because they don’t want people to be put off by their accents. Core Dumped is great and specifically put out videos explaining how they make/research their stuff and just use AI voice because they’re from South America and have an accent that they feel would be hard for people to parse.
There’s also a bunch that just pump out AI shorts with incorrect info so trying to separate the wheat from the chaff is a pain.
RETVRN to “Unregistered Hypercam 2” on the corner of a screen recording and communicating with the audience through dragging the note pad on and off screen for tech tutorials.
That’s too bad because I love South American accents. I get the reasoning but it’s too bad they resort to AI instead.
I have a friend that wants to get into voice acting and I bet he’d get into doing real overdubs in English for a fairly low rate.
Do you think there would be a good market for people to read scripts in an American accent so that people with other accents don’t have to use AI? I kinda think so because if I encounter a video with AI voice I turn it off immediately. If it’s a non-American accent I will keep listening as long as I can understand it.
Overall I think it’s better to be yourself when you make a video and not take those shortcuts but if someone has good quality research I’ll put up with some amount of AI shit.
Physics by Eugene Khutoryansky is one of my favorites.
They’re probably is, but it depends on how cheaply the video creators want to make their stuff. Or if they even care.
On top of the accent thing, there was also the reasoning that “I can sit at the beach and make my videos, and I don’t need to worry about finding silence” that was a motivation too.
Technologically different. Vocaloids work more like this and never required datacenters or any of the other ethical problems to work. AI Voice generation needed the bad shit as a once off to be made (but don’t need continuous updates forever like other uses of the same tech) but they’re here now so might as well use them.
I dont mind AI voice if it wasnt an indicator that the script was written by chatgpt. Some people are good at all the behind the camera stuff and I dont necessarily think they should have to find a voice actor for a video essay. I dunno how much data centre madness is involved in glorified text to speech, any amount is too much there. But I get the accessibility option of text to speech and get not wanting to sound like Stephen Hawking while using it.
There’s a ton of really good programming channels that use AI voice because they don’t want people to be put off by their accents. Core Dumped is great and specifically put out videos explaining how they make/research their stuff and just use AI voice because they’re from South America and have an accent that they feel would be hard for people to parse.
There’s also a bunch that just pump out AI shorts with incorrect info so trying to separate the wheat from the chaff is a pain.
RETVRN to “Unregistered Hypercam 2” on the corner of a screen recording and communicating with the audience through dragging the note pad on and off screen for tech tutorials.
What really gets my goat is that people are intentionally putting chaff in my wheat! Cut it out!
That’s too bad because I love South American accents. I get the reasoning but it’s too bad they resort to AI instead.
I have a friend that wants to get into voice acting and I bet he’d get into doing real overdubs in English for a fairly low rate.
Do you think there would be a good market for people to read scripts in an American accent so that people with other accents don’t have to use AI? I kinda think so because if I encounter a video with AI voice I turn it off immediately. If it’s a non-American accent I will keep listening as long as I can understand it.
Overall I think it’s better to be yourself when you make a video and not take those shortcuts but if someone has good quality research I’ll put up with some amount of AI shit.
Physics by Eugene Khutoryansky is one of my favorites.
They’re probably is, but it depends on how cheaply the video creators want to make their stuff. Or if they even care.
On top of the accent thing, there was also the reasoning that “I can sit at the beach and make my videos, and I don’t need to worry about finding silence” that was a motivation too.
Local offline voice models are pretty good, useful tools that will still work just fine after all the data centers are burnt to the ground.
The Elephant Graveyard special.
So you’re saying I should start a youtube history channel and use vocaloids for the voice overs?
Interesting.
Vocaloids are technically different but either way you should make your videos.
How dare you
Technologically different. Vocaloids work more like this and never required datacenters or any of the other ethical problems to work. AI Voice generation needed the bad shit as a once off to be made (but don’t need continuous updates forever like other uses of the same tech) but they’re here now so might as well use them.