Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

25 Years In Speech Technology and I still don’t talk to my computer

matthewkaras.medium.com

Post reply on HN
Print view View original

231–240 of 296 posts

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#231

Post by dheera »

heleninboodler wrote: post #136
drunkpotato wrote: post #122

Earlier quoted context omitted.

Alarms and reminders are my most frequent uses of Siri (and to entertain/frustrate the kids...). It's also pretty good for hands-free quick texts while driving and hands-free calling, though I don't do those often. Your alarm example is frustrating. Of course, you could say "set alarm for tomorrow at 6:30 am" which would work, but then you're back in the realm not of natural language commands, but formal commands tha…

Yeah, I would think it would be obvious to most humans that "set an alarm for 6:30am" never means "6:30am four days from now." It seems like it should just be interpreted as the next occurring 6:30am unless otherwise specified, but hey, I'm not a computer. Edit: also, your "tomorrow at 6:30am" example is also open to interpretation if you're saying it two minutes after midnight. I'd really like it to recognize these…

... and this is exactly the sort of thing where I feel much more comfortable looking at a GUI that clearly says the date, day, and time all in one place. It would take me 2 seconds to verify that it's exactly as I want it instead of parsing a voice assistant's response.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#232

Post by dheera »

mikestew wrote: post #143
nottorp wrote: post #135

Earlier quoted context omitted.

Ouch. I just did a quickie test asking Siri to set me a reminder for 'tomorrow at eighteen'. No dice. It set it for 8 am. It doesn't even support 24 hour time - at least in English.

"...set an alarm for tomorrow @ eighteen hundred." works just fine. At least in U. S. English, I don't think I've ever heard anyone refer to 6 p. m. as "eighteen".

Except it isn't a hundred, it's sixty, and I protest that usage. >.<. I prefer eighteen o'clock.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#233

Post by shadowgovt »

And on the flip-side of this story... "Hands-Free Coding: How I develop software using dictation and eye-tracking" (https://news.ycombinator.com/item?id=24846887)

The author notes that they only get about 50% of regular speed with this approach, and that may be a significant part of the challenge---speech can encode complex concepts into a few words (especially given context), but the actual baud rate isn't particularly impressive. Keyboard interface, where possible, seems to still win out.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#234

Post by dheera »

outworlder wrote: post #138
dheera wrote: post #5

A big problem with the assistants is that as soon as they fail at a query they seem stupid, I feel stupid, and I stop using them for a long time. Last weekend I had the following failed queries: "OK Google, what is the air quality like at Mt. Shasta today?" "OK Google, add a waypoint for the last gas station before the mountain pass" "OK Google, what percentage of people can you detect to be wearing masks on recent I…

> "OK Google, what percentage of people can you detect to be wearing masks on recent Instagram photos tagged at a location within a 10 mile radius of Mt. Shasta?" Ok, so you need the assistant to: * Already have a trained dataset of people wearing masks * Fetch ALL instagram pictures it can find * Not only detect if there's a mask in the picture, but count them * Fetch the location of Mount Shasta * Calculate a 10 mi…

> Not even things like Wolfram Alpha

Sure, but we're talking about Google here, not Wolfram. The masters at query optimization and MapReduce. I would have expected them to be able to parse this query, or at least fetch the AQI near Mt. Shasta but can't even do that.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#235

Post by blahblah4332 »

We really need to stop expecting so much from speech in technology. It simply isn't a great input method. It's loud, it lacks privacy, and for short commands it takes way too long.

I think a lot of people are counting on speech to bring us into a sort of Star Trek future.

The real game changer for input is along the lines of what the neural lace is supposed to be. Cognitive input. Silent, fast, efficient. In many cases once the technology is mature people won't even have to internally verbalize commands. Just look at a light and desire it to be dimmer.. it dims. "typing" at the speed of internalized though will also be amazing.

Every time I hear someone (including myself) tripping over "OK Google" I cringe.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#236

Post by singularity2001 »

fareesh wrote: post #58

Voice activated code snippets or command line switches would be nice. "Hey ffmpeg crop the first 15 seconds of input.mp4 and increase the volume by 15%". Modular so I can switch out the speech recognition and the NLP and fulfillment

Betty tried that: https://github.com/pickhardt/betty

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#237

Post by dheera »

ryandrake wrote: post #150
PeterisP wrote: post #137

Earlier quoted context omitted.

GUIs won over console workflows because GUIs have better discoverability and the "recall vs recognize" difference; it's mentally much easier to recognize the option you want when presented it than to recall the existence or the naming of that option. In those aspects of UX, voice interfaces have the same drawbacks as console apps when compared to a good GUI. Also, they have to work within the "bandwidth bottleneck" o…

> Also, they have to work within the "bandwidth bottleneck" of audio - just imagine a phone system that tells you all the options you have, "Press 1 for something, Press 2 for another thing..." - they are so annoying because they are slow and inherently linear They're even slower now because the brain trust decided adding voice control to the phone menu system was a great idea. So before, it said "For prescription re…

Totally. I cringe at FedEx's system.

"Welcome to FedEx. [... blah blah blah ...] Tell me what I can help you with today."

"a package"

I mean, what the hell else can you help me with today anyway?

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#238

Post by p1necone »

I think trying to replace regular GUIs and mechanical inputs in every use case is naive, it seems pretty obvious to me that clicking on a button or pressing a keyboard shortcut is pretty much always going to be faster than uttering a voice command.

But there's a huge swathe of use cases where it does make sense, and I think we should be focusing on those - situations when you can't use your hands. Voice assistant technology doesn't even need to be great for this, just good enough that you can look up unit conversions with your hands covered in bread dough, or navigate while driving or whatever.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#239

Post by Supermancho »

blahblah4332 wrote: post #235

We really need to stop expecting so much from speech in technology. It simply isn't a great input method. It's loud, it lacks privacy, and for short commands it takes way too long. I think a lot of people are counting on speech to bring us into a sort of Star Trek future. The real game changer for input is along the lines of what the neural lace is supposed to be. Cognitive input. Silent, fast, efficient. In many cas…

> It simply isn't a great input method

Loud is relative, privacy is an aimless indictment that's orthogonal, and for brevity?

Speech is a fantastic tool for communication, which includes input and output. It's part of why most large animals, for which quick communication is imperative, use it. It's imprecise, which is the problem that machines are not good at dealing with. It's was a good direction, when we used to have machines that were initially trained with speech for better accuracy, but now passive listening of devices isn't even used for that!

> The real game changer for input is along the lines of what the neural lace is supposed to be. Cognitive input. Silent, fast, efficient.

Silent sure. The human mind is rather random, highly variable between individuals and ages. I would not call it fast or efficient. Then again, speech to text is contextual cognitive input. Without drugs or intentional damage (minor) to the brain, I don't expect neural implants to be very effective, even in the next 100 years.

Re: 25 Years In Speech Technology and I still don’t talk to my computer

#240

Post by thomastjeffery »

lm28469 wrote: post #49
dheera wrote: post #5

A big problem with the assistants is that as soon as they fail at a query they seem stupid, I feel stupid, and I stop using them for a long time. Last weekend I had the following failed queries: "OK Google, what is the air quality like at Mt. Shasta today?" "OK Google, add a waypoint for the last gas station before the mountain pass" "OK Google, what percentage of people can you detect to be wearing masks on recent I…

As soon as you see them for what they are it makes perfect sense. They're basically a proxy to google search with extra dumb features on top. Every single feature that isn't google search has been hard coded (map directions, setup alarms, &c.), as soon as you get out of the hard coded cases it's blatantly obvious these things are anything but "smart" imho they're just used to give google/amazon a few more data points…

And this would be a much better situation if users could hardcode their own features and definitions.
Post reply on HN