I use them almost on a daily basis to control the lights in my apartment, it's surprisingly convenient and helps with multitasking when I'm in a hurry.
The same thing goes for weather and reminders. The best use case when you're in the middle of something, eg. cooking and just want to make a simple note or set a timer.
Does it actually work for you though? I've quickly abandoned Google assistant because it just misunderstands what I say about 2/3s of the time. A friend of mine who has all his lights controllable by voice complains about the same issue - a lot of the time, Google just doesn't understand the command so by the time you get it "right" you could have walked up to the switch 3 times.
I had Alexa controlling the lights for a while, but I found talking requires much more cognitive effort than picking up a remote, finding the right button by touch, and pushing it.
I can literally do the latter three quarters asleep, but not so much the former.
It would be far more useful to have the process almost completely automated. Lights go on when someone enters a room and go off when everyone leaves, with optional manual override.
This turns out to be a hard(ish) problem that needs better sensors and/or some form of personal ID.
Using Alexa you can set up routines, which are effectively macros triggered by a keyword, and tend to be recognised more consistently. I’ve got mine set so “bedtime” turns off the main lights and turns on dimmed side lights, and “goodnight” turns off all the lights in the house.
I switched from g-assist to Alexa because at least in my experience I found that while google was way better at random trivia, Alexa was way better at understanding the narrow set of commands she supports.
I think this is both a mic quality issue and a nlp issue. All the online comparisons compare random trivia but I almost never ask trivia because the failure rate on both is too high (that domain is way too open ended for assistants right now I think).
It does, just need to remember that they require a specific way of talking. I'm always by default talking more loudly and articulately with simple expressions, then it gets it for the first try.
(At least Google Assistant, Siri has way way more issues understanding me.)
[Disclaimer: I work for Siri; discount my enthusiasm accordingly]
For me, Siri when holding my phone is the least compelling use case. I kind of like using it for Alarm Clocks and Timers, because recognition in that domain is quite reliable, and one instruction saves multiple clicks.
Since touch navigation on a watch is less convenient than on a phone, some further use cases become more convenient with voice than with touch, e.g. asking "Is it going to rain today" before stepping out the door.
On Homepod, music is an obvious use case, which is unfortunately a rather difficult domain, because of the wide variety of media names. I use commands like "play some John Coltrane", "Shuffle play list Aggro", but also "who's playing piano on this track" (availability is a bit variable) or "what song is this". Home control is also convenient.
With Airpods, music is again the obvious use case, but I also like using them for walking directions (because you can walk without having to constantly glance on your phone). They also serve as a "poor man's CarPlay" in cars not equipped with a suitable media system (With the transparency mode on AirPod Pros, I feel that they are not an undue safety risk).
CarPlay is one of my favorite use cases, because I can navigate, listen to music, and listen and respond to messages without having to take my eyes off the road. When stuck in traffic, I also like asking what my ETA is.
Good to know.
As you work for Siri(Apple)
I want to learn and contribute in NLP fields as well.
1.Can you share your journey if possible?
2.Is Ph.D or masters necessary and how helpful the work related is in the company vs the research one does as part of Ph.D(because i have heard one doesn't get autonomy to researh one's own subjects under the supervisor but need to do what supervisor says)
3.Can you share any resource you lookup to learn and about Siri internal's working
4.How deadline works in research field ,currently working as a Front end developer deadline in my work as per ETA we can judge roughly,but how it works in research oriented field where one is unsure whether things are delivered as per requirement
1. I have pretty much a pure programming background. No previous experience with speech or machine learning when I was hired (but that was long ago).
2. The software engineers in the team have a variety of backgrounds. I think I may be the only PhD, and my subject was not relevant to the job. Plenty of people with bachelors (for visa reasons, non-US employees tend to have higher degrees). Machine learning knowledge helps, but is not strictly required. For the data scientists, on the other hand, advanced degrees are a definite plus, and so is some specialization in a relevant subject.
3. The most important thing to know about Siri's internal working is that we don't talk about Siri's internal working…
If you randomly would like to learn something to improve your chances of working at Siri, a machine learning class (e.g. Ng's and/or Hinton's Coursera classes) could definitely help.
4. There is still a lot of engineering involved, and often by the time a formal schedule is worked out, the scientific discovery part has largely been solved. Sometimes features end up not working out and have to be pushed back. What helps for Siri is that a lot of the complex functionality is server side or in updatable assets, so iteration is a possibility.
5. That's above my pay grade, really. What I learned the past few years is never to bet against deep learning being able to tackle a particular problem, but I can't shake the feeling that we'll discover limits some day.
I use Siri all the time for sending/reading text/WhatsApp messages and making calls and setting/stoping alarms and timers, and creating notes and taking pictures (Siri will open the app to where you are a click away from taking the picture. I also use Siri as a dictionary and weather forecaster.
I wish Siri was more powerful, to do things like create contacts and turn of the phone but Apple seems super conservative about what it will allow Siri to do
Sending & reading messages in public?,because apart from home automation and minor tasks,telling and hearing private infos through assistant is not that comfortable i feel.
I set alarms and reminders.
I ask how to say thing in various languages, then ask those words in that language to see if Google say the right definition. I check facts.
Fact retrieval performance varies a lot and is a good test of an assistant. Alexa is very good, because it can't rely on a screen (and it seems a lot of people work on making it good at knowledge type stuff). Google is also decent most of the time. Apple often defaults to "here's what I found" which is useless.
* Open a certain app on my watch without touching it
* Start a phone call or facetime
* Play some music —> this starts a personal radio station in apple music
* Add a reminder to my grocery list or another specific list. Or “remind me tomorrow”
* start a timer on my watch
* Set an alarm on my watch
* turn on/off smart home lights
* Some stuff with Siri Shortcuts. I use this mostly for time tracking timers, starting and stopping. But sky’s the limit here, you can start any multi step automation from siri if you have a use case that doesn’t require input
Bixby responds to almost anything, so it is useful to find my phone. It can answer a question about my calendar, am I available at 9 am tomorrow, but it useless for setting an appointment. ‘Create an appointment with Bill tomorrow at 9 am’, will create an all day appointment with a location 9 am.
Alexa doesn’t integrate with my calendar at all. (incompatible)
I use it for ad hoc hands free navigation while driving. It works 60% of the time and I don't have to intervene with my hands.
60% is sort of like a gamble with no downside. Do I have to stop the car or can I just set the destination while driving? I would have to stop the car anyway were it not for voice activated commands.
I use my Google home for timers while cooking, turn on/off lamps that have awkward to reach switch's, Christmas tree lights when the time arrives. My biggest usability issue hey/ok Google is a mouthful and does not roll off the tongue like Siri/Alexa.
I use google assistant multiple times a day. play pandora or youtube playlists on various speakers throughout my house, set reminders, add things to my shopping list or todo list, check the weather, set alarms, navigation if I'm driving, random google queries
