Arguments that WIMPs' (windows, icons, menus, pointers) status as the dominant user interface paradigm are under challenge like never before just got a little stronger, after Google announced that a new beta of its Chrome browser will adopt the Web Speech API. The API is a W3C initiative that makes it possible for browsers to …
" A voice-enabled Android browser would take the mobile OS's voice-recognition capabilities beyond its current Voice Actions abilities."
Android has had a fully functioning voice search for some time. It also has continuous voice dictation in stock Android too and has done since Jelly Bean. This is on top of creating calendar appointments or finding out the mass of the sun etc.
In my experience testing the two it works a lot better than Siri also.
But Siri is the least used program on my iPhone since I asked what the latest Championship League tables were and it gave me the biography of Adolf Hitler.
These voice enables are a pure gimmick with very limited appeal.
Also when in company you look a pr*ck talking at your phone or computer, then getting angry because it just doesn't understand you.
Doesn't even know who Pussy Galore is.
I find it silly but my wife uses siri all the time to add reminders and it works remarkably well, for that use case I'm grudgingly forced to realise it is quicker to do it that way. Though it's the natural language processing which is really the time-saver.
I use siri reminders all the time, also to find out what the weather is going to be. Searching is pretty naff. Calendar appointments are tricky to get working. just need "Siri turn the wifi on/off", or "Siri turn the data on/off"
not possible on non jailbroken devices.
Well you have to admit, the biogrphy of Adolph Hitler is a lot more interesting.
Google voice search gave me a link to the Wikipedia entry for "Who is Pussy Galore". Hate to think what I'd get if it mistook Who for What.
Whether or not it works better than Siri is irrelevant. Until voice activation can work with 100% accuracy 100% of the time for 100% of its users it will remain nothing more than a curiosity and a toy ... on any device or platform.
So you've never had to use the backspace key or click on the cancel and/or undo button? Voice will win whenever and wherever it's more convenient than keyboard, mouse, and/or touch.
100% accuracy ....
100% accuracy? Hell, even a human stenographer isn't going to give you that.
No, it just needs to be "good enough".
100% Accuracy is not needed
The non WIMP interface only needs to be as good as an administrative assistant. After a few months on the job a good assistant should understand most verbal orders and get most face to face commands right 99% of the time. A highly competent assistant should have your daily routine down pat and recognize out of the ordinary (stressful conditions) where patterns break down and unknown unknowns dominate potential errors. Under stressful conditions (all nighters) your computer should present you with several possible interpretations of your commands.
Sometimes a dumber assistant is needed. Like General Grants aide (rumored to be the dumbest 2nd in the Union Army). General Grant knew that if his aide decipher the intent of the string of words in a written order then it was ready to be sent.
"Until voice activation can work with 100% accuracy 100% of the time for 100% of its users"
That's WAAAAY above even the best human secretary, let alone one who's tired / hung over / distracted playing solitaire
Just remember if your administrative assistant (nee secretary) took dictation at a 99% accuracy, any executive worth their salt would just as soon fire them. Given that most voice recognition is not that near, as said in a previous post, it remains a nice fun tool. You will always be asked a confirming question (did you mean...?) and go from there.
Yes, great strides have been made, but it isn't "production quality" for the most part. I wouldn't bet my life on the recognition the first time. Even us humans don't get it right all the time and need to ask questions.
Life goes on.
Am I the only one...
...Who hates talking to a device? I don't care how good the speech recognition is, outside of a few, rare, very specific circumstances, I would feel a right wally talking to my technology. Give me a mouse and keyboard, or a touchscreen where appropriate, but don't expect me to talk at it. I can't even stand speech driven menu systems companies use for incoming phone calls, I would much rather press a button on my keypad.
Re: Am I the only one...
You're certainly not alone. If it worked well though, it could be a handy mitigation for the "wot no keyboard?" problem on mobe/tab/phab thingies :D
Re: Am I the only one...
I hate talking to people let alone devices.
This isn't just being a bitter old man, part of my job is to tell a bunch of people what to do, (basic stuff like move that over there, point it that way, turn it on, now adjust it) and there is not a single instance where it wouldn't be far quicker, more effective, less stressful and just make me feel far more happy to just do it myself.
Try to do any complex task but only thru giving verbal instructions to another person and even if that other person is extremely intelligent, well trained, and highly motivated to help you will still very quickly curse the very air you breath.
Yes a voice comand device is useful for basic symple tasks or list, and can occasionally settle drunken debates, but real actual productive work its always simplier to just shut up and do t yir self.
Re: Am I the only one...
I don't know about talking to them, but I spend a lot of time berating devices. Will an Android phone ever be able to actually Fuck Off when told?
Re: Am I the only one...
Only if it sprouts legs.
Haha, yes indeed. I hate talking to them but I am more than happy to shout at them.
...just what I need when I'm on public transport or simply out about town - more people talking to invisible friends. Because I currently like nothing more than being forced to listen to an idiot "talking" to his/her iphone while holding it a foot away from his/her mouth. And if I'm *really* lucky, they've turned on the speaker so they don't have to use the ear-pieces - so I get to hear *both* side of the loud conversation.
And now this madness will spread to android devices. Oh joy, oh rapture.
And for the impaired... yes, that was sarcasm.
Re: Oh good...
"Oh joy, oh rapture."
Oh no, "app rap"
Speak into the flowerpot
"I need a truck", or.....?
Using Voice and Gestures on PCs is Necessary for Speed during Video Conferencing
There are one person use cases where a handheld device is useless without voice commands. I'm not talking about those.
The use case where several people are collaborating by modifying documents in real time is what I'm talking' 'bout. In that case several people sitting at desks and one or more leaders pacing in front of large monitors are trying to polish a product or agree on descriptons of features of some product.
Voice commands and confirming gestures that can be implemented the first time (with few exceptions) using non WIMP and non touch screen interfaces would be highly efficient in the above use case.
Re: Using Voice and Gestures on PCs is Necessary for Speed during Video Conferencing
I see where you are going. However I'm not sure how that would work. Voice may be expressive and precise, however editing it is fairly hard. You would need to create languages in which you can express ways to correct what you have just said.
Voice is also not necessarily fast. Optimized keyboards can reach the same channel capacity, and the upper limit is the roughly 15 bits per second you can output anyhow. (That's 15 bits of information, not 15 bits of data)
An efficient voice command system would probably deviate from normal speech. You would use special mnemonics and utterances to get your commands across. Natural language is just not made for such things.
All I Want Is A Return To Voice Dialling
With all this talk of speech recognition being the coming thing, why is it that mobiles have gone backwards with this technology? My old Nokia does voice calling perfectly well but can I get a modern phone, smart or otherwise to do what a nearly 8 year old phone will do, in the modern words, 'the computer says no'. If it cannot do proper hands free via blue tooth why bother with other rubbish? My recent Nokia Asha 300 carrier down grade sits back unloved in its box waiting for recycling.
[PS If anyone knows of a phone that does voice dialling as well as a Nokia 6230i, uses blue tooth and fits in a shirt pocket please tell me - large expensive junk need not apply.]
Chrome has had the Web Speech API for over 2 years.
Chrome? no thanks!
SRWare Iron? go on then!
If they can get it working for my Geordie and Mackem mates I will buy Google a pint.
Sadly though, I suspect they will be arguing with their devices for a long time to come.
I tried Siri.
All it ever seemed to do was offer me a web search, except for "When is the next train from Glasgow to Edinburgh?" which gave me "I do not have an event for next train from Glasgow to Edinburgh". It might be OK for hipsters looking for the nearest moustache waxing shop in San Francisco, but for real life in the UK it seems pointless.
So not only will Google know my interests and what I buy, they will soon also know what my voice sounds like.
Worse than that - with these moustache games, will it be able to get video of you too?
It knows the answer
I asked my Nexus 4 what the answer to life, the universe and everything is, it knows...
Yes, very cool stuff. We will soon have Star Trek houses with personal droids (running Android!) as valets. More immediately, if this becomes the primary interface for the masses, it seems that our workplaces will become something like a telemarketer sweatshop with everyone talking all the time rather than pounding on keyboards and quietly surfing the net. One obvious opportunity will be in developing really good noise cancelling office equipment.
The main problem: it's annoying to others
After all voice is how we communicate with peers. When you talk to your computer you will not only occupy the acoustic channel in your area so others won't be able to talk while you talk, but people will also instinctively listen to you talking to the computer.
So even if it would work perfectly, it would at best be rude.
apple have already patented all forms of speech.
- Vid Antarctic ice THICKER than first feared – penguin-bot boffins
- Hi-torque tank engines: EXTREME car hacking with The Register
- Review What's MISSING on Amazon Fire Phone... and why it WON'T set the world alight
- Product round-up Trousers down for six of the best affordable Androids
- Antique Code Show World of Warcraft then and now: From Orcs and Humans to Warlords of Draenor