Any progammers looking for a killer app to develop? How about a voice enabled forum?

One of the most powerful, exciting, and engrossing experiences on the Internet is the Forum. The first great Internet forums were the Usenet newsgroups. Usenet is still a powerful force, but many different types of forums are also very popular (such as message boards like Vbulliten and XMBforum).

I love forums. Love em love em love em. My web site has a great one. But I think the time has come for an important and breakthrough evolutionary advance in Internet Forums. I will describe the idea to you, and I hope you will take the idea and run with it. Create this product. Develop it. Make millions of dollars. I offer this idea to you-because I am not a programmer-and I want this to be made.

The idea is pretty simple. Any fairly savvy application developer with some skills in MySQL should be able to do this. Whoever does it right will make a lot of money.

I want someone to simply improve upon the Internet forum technology.

The amazing power of a forum is that you can walk into a forum and have a conversation. You can ask a question and get insightful, funny, and interesting answers. You can then reply and have a dialogue. You can do all this at your own pace, your own convenience. You can ignore people who don’t interest you and you can be as patient as you want in replying. It is a wonderful thing. And the time has come to liberate this experience from the necessity of a keyboard and full powered computer. The technology is here, we just need to take advantage of it.

This is what I want.

I want to be able at any time of day, whether driving my car, or doing dishes, or hiking through the woods -to ask a question of the world. I don’t want to have to find a computer, go to the website or newsreader, and type up my question. I want to just be able to say outloud, “What is the fastest any human has ever travelled?” or “What the heck does a county commisioner do?”…and I want my question, an mp3 recording of my actual voice, immediately posted in a forum. And then I want to receive responses-but not responses i have to go sit down and read. I want to hear the responses, in the voices of those listening to my question.

Imagine, you could be driving in your car…and you have a question. You simply say out loud, "Post this question: Can anyone give me more details about the Xbox2? Or “Can anyone tell me real quick how to make tomatoe soup?” Or while reading a history book…you could say outloud, "Post this question: “Was Wyatt Earp a real person?” Whatever you are interested in. For whatever reason you communicate with people in Forums, you could do it-but simply with your voice-at any time anywhere.

After you post a question, you might hear a little beep in a few minutes letting you know you have responses. Then you could just say out loud “Play replies” and the replies would start to play-not in a computer voice but in the actual voice of the people answering you-because they would be using the same technology.

With this type of voice message board-you could have engaging, interesting, convenient, and fairly natural conversations over the Internet. The experience could be liberated from sitting at a computer-and it would be enhanced by being able to hear human voices-rather than reading text. You could have all the fun and convenience of natural conversation-but enhanced through the reach and expertise and convenience available over the Internet.

That is the concept. It may seem daunting, but the technology already exists and is available. All this idea requires is a MySQL based message board (such as the XMB Forum) and a pretty basic client side application. The SQL database could store all the voice messages in mp3 format. The client side application would run on a pocketpc phone or palm phone.All this application would need to do is

  1. interface with the MySQL database (the message board). Be able to upload and download the mp3 files in the appropriate thread-and play them on demand.

  2. encode into mp3 format as people recorded their posts.

  3. This part is optional, but would be the ultimate addition: be able to recognize some simple voice commands such as “Record” “Post” “Replay” “Skip” “Next” “Back”.

Thats it. If you can program or hack MySQL and make pocketpc Apps, then you can make this amazing voice enabled internet forum.

Limitations:
Bandwidth for moble internet access is an obvious limitation. But remember, voice forums will be just as fun and convenient on desktop computers and where high bandwidth is available.

Webstorage. Voice messages will be much bigger. The MySQL database will grow fast. But hey, it will be so fun, I think users will be willing to pay the message board host a little to help cover storage costs. Another alternative is to delete or archive all messages after 3 days-limiting discussions to fairly current threads.

Ok thats it. I know this will arrive someday. But why not sooner rather than later? Any programmers out there interested in a little project like this? If so, go for it. Make millions. I maintain no copywrites on this idea. But if you are going to run with this idea, please just let me know so I can use the forum on my website.

The best way for this to be developed is through an open source project. Anyone who wants to collaborate on this project with other programmers feel free to post on my website (The Wisdom Project Science and Technology Forum) to make contacts. Or just go for it if you already know how to do something like this. Good luck. I look forward to hearing all about it!

There’s no programming challenge here. The real problem is one of infrastructure and hardware support, and to get that, you need a metric shitload of money. I don’t see that coming any time soon.

I disagree with ultrafilter. You’d at the very least need to do some voice to text translation so that you could slap headers on the files in the database. Unless you really think that the world is going to sit around listening to random voice snippets looking for a question they’re interested in answering. Probably you’d have to go the whole natural language route in order to translate a voice query into something remotely indexable. At which point, why bother submitting to a forum? Feed the results of the indexing into Google and see what you get.

Voice forums have a lot of disadvantages over text forums. Try getting a specific bit of information out of a lengthy voice message on your answering machine and you’ll see what I mean.

I’m not quite clear on how you’d extract the millions of dollars from this scheme. I think it would be the other way around.

At this point, voice translation software exists, and it’s good enough that you could translate speech to text. And that alone is good enough for posting on a messageboard.

I agree with ultra filter. The solutions to create what you want already exist, but that’s not enough.

There are several problems. Off the top of my head:

Bandwidth and storage space, as well as the hardware requirements. You think this board chugs along now (those poor little hamsters!), try having several hundred people dropping several hundred kbytes worth of audio data on the servers AT THE SAME TIME. My god, the hamsters would strike for sure!

Additionally there’s the problem of the moderators. They can’t scan through the text or run searches, they’d have to edit audio files. Sounds time consuming.

And also brings up another point, searching and scanning. How do you search thousands of audio files for the words: “star trek vs starwars” efficently?

Just wait a couple years until the semantic web gets organized and going strong.

[Tangential nerd moment] For those who don’t know the semantic web is a effort under W3C to move the web out of isolated information, into information having meaning and relationships to other data. The main effort is so that computers can communicate ideas without needing specific hand-coded definition conversion. But the side effect is that all data will be encoded with meaning so that questions can be answered with an Inteligent agent searching your query, and all related branches until it finds the answer and returns it to you all organized and shiny[/tangential nerd moment]

Hehe got distracted there , but basically I believe the ‘question and answer forum’ will be obsolete before the voice technology gets implemented.

Infrastructure:
As far as storage space is concerned, members could simply be asked to subscribe to cover the cost of storing their voice messages. $5 a month per message board member would provide a ton of space.

“Natural Language translation” is not necessary. The basic filtering of posts could be done by the poster. You could have different forums for different types of discussions. A general questions, IMHO, Café society, and MPSIMS would be adequate. The posts don’t need to be searchable. You could simply start listening to a post and if the post didn’t interest you, you could say “skip” or “next”. Perhaps an etiquette could be established where posters said a brief title to their post at the beginning of their recording.

The point of this system is not simply to be able to use speech to get “right answers” to questions. It is about using speech to reach the minds and knowledge of other people. I don’t want a google search, I want the equivalent of walking into a room of people and asking a question. I want to be able to ask-what are the good and bad of the Medicare bill? Or “Why do bad things happen to good people?” Or “Are we wasting money on the Mars Rover project?” And then I want a discussion to begin-a natural discussion with real voices-simply moderated by the convenience and flexibility of the Internet.

The ability to search the archive of discussions would be cool, but not necessary. If we could simply recreate the experience of a café where people get together and discuss things, that would be amazing. It would simply be a café we could enter and talk to people-from anywhere at any time.

As far as moderators, they are not necessary. I believe members could moderate by rating messages like on Slashdot.com and reporting flames. Moderators would only need to listen to reported posts to see if they violated posting rules.

Thanks for all the input. By the way, if anyone thinks they could easily build such a MySQL database and application, I will pay you-but my funds will max out at around $400.

Thanks.
Matt

This is the biggest problem with the idea, IMO. I’d much rather read a thread than have it read to me, which would take MUCH longer.

If the voice messages were stored as low bit rate mono MP3s, they would take very little space. HD space wouldn’t likely be an issue. As an expiriment, I recorded a minute of text (from the OP) as WAV then compressed it into 48k mono mp3s and Ogg Vorbis. The MP3 was 352k and the OGG was 343k. Both sounded excellent.

That means with only 20gig you could store almost a thousand hours of speech.

No one’s going to pay for this until the infrastructure is already in place.

Sorry, don’t mean to be a killjoy, but the idea is deeply flawed. The problems aren’t technological (although the bandwidth costs would be horrendous), they are to do with useability. I can read and scan text many times faster than I can talk or listen. A text based forum would be dozens of times quicker to use. How would you browse a voice forum? Only by listening to a list of all the topics, which would take ages. People simply wouldn’t use it.

A better idea might be a forum linked to a teamspeak server. Clicking on a topic could take you to a TS room where people were talking about it.

Frankly, I don’t want anyone to hear the grating, nasal way my voice comes out when recorded.

I imagine it comes out that way in reality, too, I just like to think it doesn’t.

Absolutely. Why is this a killer app? I can’t browse audio files at work. Many companies block multimedia files. There goes some of your userbase. I can read faster than I can listen; I have no interest in wading through each individual audio file, each of which is larger than a text block. Not everyone is on broadband, especially outside the USA; pity poor dial-up users.

Not a bad idea, but in no way a killer app. If that was true, audiobooks would be outselling normal books, wouldn’t they?

Here is another bit of evidence that this idea would never fly:

Back in the 1980s I used to receive 20-30 voice mails a day. Then email came along. I now receive 40-50 emails a day (not including SDMB subscription notices :slight_smile: )

Email is FAR more efficient than voice mail ever was. I used to spend hours handling voice mail. Now I can handle twice as many “messages” containing ten times the information, in one quarter the time. That’s the efficiency of the printed word over the spoken word.

This forum has voice and video. Not quite what the OP is looking for, but …

This post covers exactly the two points I was going to make: audio is not preferable to text, as exemplified by the fact that books are almost exclusively still printed rather than recorded.

Moreover, there are some things for which audio is almost completely useless. Try explaining math, physics or computer programming to someone over the phone. Anything that requires equations or computer code with a specific syntax is a bitch and a half to discuss verbally.

This idea reminded me of the bad Michael Crichton book “Disclosure” (I think that’s the title – they made it into a bad movie with Demi Moore and Michael Douglas). The story takes place in a software company which has developed an amazing application: using virtual reality, you can organize thousands or millions of files by storing them in virtual filing cabinets. When you want to find a particular file, you don your virtual reality goggles, walk down a virtual corridor, find the right virtual drawer, find the right virtual file folder, and then read the information written on a piece of virtual paper. See, it’s just like filing things in real-world filing cabinets, but, um, the rooms are bigger! Wow!

As pleasing as the mental picture of asking the Internet a question like Jean-Luc Picard is, it’s just not a natural way to interface that many people and that many different kinds of data. A visual interface works much better than an audio one.

Ahh so not one person sees why this service would be a compelling and rich experience?

The point of a voice based forum is not to provide a replacement for text based forums. Obviously voice is far less efficient for the raw task of disseminating information. You might say the point is to provide an alternative to live discussion groups, that is, the types of discussion groups where you physically go somewhere and sit in a room with real live people and talk to them.

Perhaps many people here never use such discussion groups, because they cannot stand the inefficiency of speaking aloud to real people. But I and millions of people still enjoy the rich experience of live conversation, despite its inefficiency.

The voice forum app that I described would enable people to enjoy the rich experience of having a spoken exchange-yet without the obligation to set aside time, travel. Best of all it would give users the ability to skip any people who bore or annoy them. The voice forum I am advocating would enable people to enjoy a discussion forum with people from all over the world in a convenient and pretty natural manner.

Another incredible advantage a voice based forum will have over a text based forum-is you will not need to sit at a computer to participate in it. You could listen to posts, reply, skip around from forum to forum-while driving your car, doing the dishes, or any other type of task you might do while say listening to the radio.

How would it enable you to skip users who bore or annoy you? You’d have to listen to your message to work that out, by that time its too late.

How would you control the forum while doing other tasks?

Its a neat idea, but there is a huge, huge gap between that and actually designing something workable.

BAAAAD idea. Never work. Shame on you for dreaming.

Think of this:

I can read the news-papers a lot faster than waiting for those new-fangled “news-announcers” to say their piece. This whole “radio” thing will never work out. People simply won’t use it.

Same deal for your plan, ready. :wink:

Sounds to me you are describing something more like voice chat, but with a record and play later feature. That may be a better angle to come at the problem than as a messageboard.

Message boards are not realtime, they are more akin to store-and-forward systems, or the mail. Voice is inherently realtime, and thus better suited for a chatroom type of environment, where conversations take place then and there, with whoever happens to be there.

As to the technology for voice control, it is already here. As long as you keep to a vocabulary of maybe 20-30 words to control your system. The small vocabulary shouldn’t be a problem if you are simply using it to control the feature of the system, “back”, “forward”,“next thread” etc.
See Vocera for a running system using limited vocabulary voice controls. (I have seen and used their system in demos. It is waaay cool, but pricey)

Biggest obstacle I see is indexing the recordings for some sort of organization and searching ability. Not sure what you’d do there. Speech to text ain’t going to cut it here I don’t think. And you really don’t want your server doing that processing, limiting what kind of devices you can run it on.

How about a text-to-speech/speech-to-text app for navigating regular messageboards?

This is almost as silly as having video with my mobile phone :stuck_out_tongue: