• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

RadioGPT, AI hosts and the future of human on-air talent

Two developments this week seem to paint an ominous picture for the future of live air talent.

Spotify introduces its personalized AI DJ:
https://youtu.be/ok-aNnc0Dko

Futuri introduces RadioGPT - AI hosts for conventional radio:

I am sure broadcasters won't use this to replace paid human talent, though.
 
"ALPHA MEDIA and ROGERS SPORTS & MEDIA have signed up as beta partners for RadioGPT."

Wouldn't it be nice to have heard a sample or two of this as a part of this story?
 
Popular music has been largely synthesized for 40 years. That hasn't stopped people from listening.

In a way you can trace the start of AI to the early days of multi-track recording in the 1950s. Les Paul's How High The Moon.

AI has a basis in something, and that something is usually a human voice. You also need humans to program the AI.
 
big disagree with this premise. Spotify rolled out "interactive podcasts" and the tech industry lauded it as the future of radio, only to have Spotify dump half their podcast division a year later. So I would disregard that story entirely...it's tech boosterim

this is a gimmick. I highly doubt Spotify users want a generic VO interrupting their playlists.

If the radio industry is dumb enough to use this in place of talent (even out of market voice trackers) they'll lose their only advantage over streaming: the personal touch of a great jock
 
big disagree with this premise. Spotify rolled out "interactive podcasts" and the tech industry lauded it as the future of radio, only to have Spotify dump half their podcast division a year later. So I would disregard that story entirely...it's tech boosterim

this is a gimmick. I highly doubt Spotify users want a generic VO interrupting their playlists.

If the radio industry is dumb enough to use this in place of talent (even out of market voice trackers) they'll lose their only advantage over streaming: the personal touch of a great jock
Exactly. I've been very engaged by keeping up with recent AI developments, including the recent partnership between a version of ChatGPT and Microsoft for Bing.

In many ways AI has become a double edged sword, and shouldn't be confused with voice synthesis. Two completely different forms of tech. Most chatbot's today use the public Internet as a memory store. Neither radio nor TV would benefit from allowing a chatbot coupled with some form of voice synthesis to talk about a subject without significant filters/guardrails. And like anything else; if the public Internet is the only reference material for the bot to refer-to, there's at least a 50% chance what the AI responds with is wrong.
 
If the radio industry is dumb enough to use this in place of talent (even out of market voice trackers) they'll lose their only advantage over streaming: the personal touch of a great jock

Keep in mind that in most music formats, the DJ has no personal involvement in the music. It's been that way for 50 years.
 
Exactly. I've been very engaged by keeping up with recent AI developments, including the recent partnership between a version of ChatGPT and Microsoft for Bing.

In many ways AI has become a double edged sword, and shouldn't be confused with voice synthesis. Two completely different forms of tech. Most chatbot's today use the public Internet as a memory store. Neither radio nor TV would benefit from allowing a chatbot coupled with some form of voice synthesis to talk about a subject without significant filters/guardrails. And like anything else; if the public Internet is the only reference material for the bot to refer-to, there's at least a 50% chance what the AI responds with is wrong.
The internet has taken us full-circle. Over time, social media has taught people to argue and insult each other. It's social media's way of keeping users "engaged." Now AI is aggregating those two decades of rude, insulting rhetoric and regurgitating it on demand. The recent Bing episode with a NY Times reporter is a good example.
 
Now AI is aggregating those two decades of rude, insulting rhetoric and regurgitating it on demand. The recent Bing episode with a NY Times reporter is a good example.
Great recent example. But ultimately what becomes of nothing more than a predictive speech chatbot, could turn into a religion, or another QAnon for some folks who can't differentiate a series of algorithms, and a thinking, sentient being.
 
All the comments I saw below the vid were quite positive. One commenter said, apparently with glee, "this is a radio killer."

I think that anyone who doesn't see this as an eventual threat to DJs and radio just isn't thinking it through enough.

Of course, radio will be around for a few more decades because it's free.

The personalization part in the YT example is almost like something out of the movie Minority Report.
 
Great recent example. But ultimately what becomes of nothing more than a predictive speech chatbot, could turn into a religion, or another QAnon for some folks who can't differentiate a series of algorithms, and a thinking, sentient being.
(Color highlight mine)

That reminds me that for every genius, there is a person with an IQ of 60 (If you believe we live in a strict bell curve... more if you think environmental factors create more sub-80 IQs than those over 120).

Most people outgrow talking dolls and figures before they get beyond kindergarten. Except with appliances and cars: "turn left at the next intersection..."
 
Popular music has been largely synthesized for 40 years. That hasn't stopped people from listening.

In a way you can trace the start of AI to the early days of multi-track recording in the 1950s. Les Paul's How High The Moon.

AI has a basis in something, and that something is usually a human voice. You also need humans to program the AI.
Synthetic voices are not exactly new. Roger Ebert got a CereProc computer that allowed him to “speak” again a few years after losing his lower jaw. It was composed of recorded material over the years.

The question here is if AI-led synthetic voices can avoid the uncanny valley.
 
No visual cues to try and mask it's a 'computer' - all has to be done via inflection.

Which might be why that this will eventually succeed.
But what would it say? It's simple to type what amounts to a script into a synthesized voice algorithm and have it repeat the text back in some monotonal way. What sort of return on investment would there be to have someone script out dialog, only to be repeated by a synthesized voice over radio? Text to speech doesn't translate inflection, tone, emphasis, or if a human smiles while they talk. Any of that would require much more specific coding for every sentence spoken by a synthetic personality. Something like that would be well out of the financial benefit for any radio station or group.
 
(Color highlight mine)

That reminds me that for every genius, there is a person with an IQ of 60 (If you believe we live in a strict bell curve... more if you think environmental factors create more sub-80 IQs than those over 120).

Most people outgrow talking dolls and figures before they get beyond kindergarten. Except with appliances and cars: "turn left at the next intersection..."
When the car voice says "stop and let me out!"........
 
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.


Back
Top Bottom