-
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
Computer (voice) as Host/"Talent"
- Thread starter kfbkfb
- Start date
I'm ok with a computer sounding voice doing a voiceover during the instrumental beginning or ending of a song giving the song title, artist and maybe album the song is on, I'm just interested in the info, and an easy to understand voice.
(perhaps the estate of Dick Tufeld - the "robotic sounding" voice of the Lost in Space robot in the 1960s TV show, the 1998 movie and the 2003/04 pilot - could be convinced to license his voice for radio/song intros/outros)
Kirk Bayne
(perhaps the estate of Dick Tufeld - the "robotic sounding" voice of the Lost in Space robot in the 1960s TV show, the 1998 movie and the 2003/04 pilot - could be convinced to license his voice for radio/song intros/outros)
Kirk Bayne
There were 116,000 radio industry employees in 1999, according to the BLS. According to the BLS today there are just over 66K radio industry employees, and that number is dropping.
Technology has made this decline in radio employment possible, and that technology includes "AI", which is the "theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages." The computers don't need to have actual intelligence. They just are designed to simulate it.
"AI" is already in the news media in some form. WA Post, NY Times, Reuters, and the BBC use it.
whatsnewinpublishing.com
The idea that AI can't or won't ever be combined with voice synthesis in the future is laughable.
It's hard to believe that anyone here would think that any such advancements in voice synthesis will never be used in radio. Radio has always embraced new technologies.
Technology has made this decline in radio employment possible, and that technology includes "AI", which is the "theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages." The computers don't need to have actual intelligence. They just are designed to simulate it.
"AI" is already in the news media in some form. WA Post, NY Times, Reuters, and the BBC use it.
Artificial Intelligence in media: Automated content opportunities and risks
What are the use cases and challenges of using automated data analytics in journalism? Information flows faster every day. As a reader that can easily
whatsnewinpublishing.com
The idea that AI can't or won't ever be combined with voice synthesis in the future is laughable.
It's hard to believe that anyone here would think that any such advancements in voice synthesis will never be used in radio. Radio has always embraced new technologies.
They use it for their online assets, not for broadcast.
Maybe laughable to someone who doesn't understand how radio as a business works, or that voice synthesis isn't AI....
No it hasn't. That's just silly. And certainly not in the twenty -or twenty first centuries. Radio uses only the highest tech it can afford. Being able to automate a radio station with a $1,000.00 PC, is just one example.
Recent work with radio automation in the cloud is a great example. There are all sorts of benefits to moving automation to the cloud, but most radio stations aren't interested because it involves a subscription charge.
Oh sure, hearing the call sign is so important to your average listener.
I understand more of how radio as a business works than you think I do. My job put the jobs of probably tens of others, at the very least, out of work. Then, 15 years later, the jobs of others made my radio job redundant, and I was out of work. All thanks to advancement and adoption of technology -- storage technology, advancement in programming technology, and advancements in the use of internet technology.
The fact that a station owner -- whether individual or corporate -- can automate a station with just a $1K PC is an example of this advancement in technology being adopted by radio. A $1K computer today can do a lot more than a $1K AT PC computer could do in the 1990s when the company I worked for developed and sold its first MOHD automation systems, when one Seagate gigabyte drive was very expensive and the size of a small to medium sized brick.
And accounting for inflation, that $1K AT PC in the early 1990s would have cost maybe $2,000 - $3,000 today.
I just don't see radio remaining static, and I don't think that AI / voice synthesis and other forms of synthesis of human vocals and images is just going to go unnoticed by radio, as well as other entertainment platforms. Where they see a way to make money off of it and cut costs at the same time, they will undoubtedly do so.
Even if and when all radio goes online. As we both know, the definition of "radio" has changed. It's not just OTA now.
And most of those stations that you said would not automate in the cloud because of a subscription charge -- OK, I get that, a lot of owners don't have tons of money to throw around when the old tech is still working and keeping the programming on the air.
But will all those individual stations even be in existence 20 years from now? Or will many of them instead be absorbed into bigger operations that have the ability to further automate and use new technology? And as technology advances, the costs drop. You should know that. There's that aspect of it, too.
Radio has never been a ground-breaker when it comes to technology. Instead, radio jumps on the coattails of other industries who spend the money in R&D, and only after the technology becomes cost effective. The days of dropping $20K on an audio console are over. Since covid, you can run a station live from your closet using an iPad. None of which is groundbreaking tech. It's much more cost effective to pay some guy like Paul way up in the tundra of Alaska to cut a few liners and voicetracks, rather than some mechanical-sounding synthetic voice with no inflection. If anything good came out of the pandemic; it made it possible for lots of cheap consumer products to run radio facilities without studios, offices, or expensive broadcast-specific gear. Besides, it's just radio we're talking about here. Audio-only, nothing fancy.
And I've personally lost count of how I've implemented automation in broadcast facilities. In fact, I invented Centralcasting for TV, which is very widely used today. As tech goes, trying to save money using synthetic voices is, by comparison, a drop in a swimming pool, and would probably have more deleterious effects rather than just paying 'Permafrost Paul' $50 to cut some liners.
I wouldn't say PC's coming down in price is advancement of technology. More like as the technology becomes more commonplace, the price naturally comes down. That, and software developers figured out a way of doing more with less. I remember installing some of the first PC-based automation systems. There were file servers, RAID storage, workstations, expensive sound cards, network gear, and an even more expensive ongoing support agreement. Systems like that were out of the reach for small to medium market stations. Eventually, there was no need for servers, networks, and alike. As the price of PC's went down, could use $30 Soundblaster cards, so moving to PC automation became an inexpensive requirement.
But that's the barrier with synthesized voices, natural human sounds. That's why most video games with realistic graphics use human voice actors to record words and lines.
This is true. So if I can pay a contractor like 'Frozen Wilderness Paul' to record my weeks worth of liners and voicetracks for a week, why would I spend money to interface some syntho-voice system into my automation? Paul isn't an employee, where the real expenses are. I'll never recover the capital cost of that gear, because it would be obsolete in four, or less years, let alone the cost of support. Even at a couple hundred dollars per week to pay actual human voices with no benefits being paid, humans win.
Here's an example of one of the several commercials I hear on SiriusXM using "AI" text-to-speech synthesis. They all have the same human voice at the end giving out the phone number.
Yeaaaahhhhh, not quite ready for prime time just yet. Lol. Not sure who approves commercial content for air at Sirius, but any PD I've ever worked with would flatly refuse to air this - But they would offer to have one of their production staff voice it...for a fee, of course.
And if it's just the required disclaimer, who cares what it sounds like? How many of these disclaimers have we heard where someone cuts every microsecond between words, let alone syllables, to make it fit within the time window? Even that sounds mostly unintelligible, but it meets the legal requirement. As if someone will say: Gee, that disclaimer sounds like a bad robot. No way I'm going to buy that product or service! No, actual listeners don't care.
AI is profoundly different than voice synthesis (as used by computers) and is absurdly more difficult. The cost differences would be humongous between the two technologies. And, they serve a very different purpose. Broadcast radio will be long gone before anything approaching true AI is useful (or can be afforded by a declining industry).
Having started in military radio at the tender age of 17 I'm on radio's side but the reality is it will go away just like the horse and buggy and the dial telephone.
Having started in military radio at the tender age of 17 I'm on radio's side but the reality is it will go away just like the horse and buggy and the dial telephone.
The insertion of the phone number is just as bad as with a human voice!
Actual listeners (or geeks) tend to crucify a badly done commercial.
Totally agree. I was listening to a podcast just yesterday, where an expert in AI development and research was talking about advances and purpose in AI development. True AI, not just voice synthesis, is very purpose-built in development. An example the guest used was having AI identify a house cat when seen in a photo. He said that one can't just program where if it has whiskers X-long, a triangular shaped nose, and eyes that look a certain way, it's identified as a cat. Modern AI uses literally billions of lines of code to learn examples, down to a pixel, of what is identified as what.
Here are modern examples of rudimentary AI robots, which is pretty amazing, but not humanoid enough to duplicate expression that would actually fool someone:
Or they take them about as seriously as they take the computer-voiced telemarketing/scam phone calls they get: "Hell o. This is Fed...Ex. Youhaveunclaimedpackageswaitingforyou.... ..... ... Please pressonetospeaktorePREsnTAT....ive. (click).
None of these are true AI. The only cost effective use for robots (which have to be programmed by humans as computers don't learn by themselves) are repetitive tasks. Other uses include 'dangerous duty' such as bomb exploration or radioactive environment detection but these are highly developed robots and, again, have to be programmed by humans. There is just no effective means of having a machine 'learn' on the job.
I also don't see a future for any machine that a 3 year old human can already perform (and, older than that, much, much less expensively).
Not yet.
I'm guessing that no one alive today will ever witness that. And, unless we get our world wide priorities straight in the next few years there will be no listeners for any radio stations still existing.
- Status