I put an Echo Dot in my son’s room a couple months after he was born to play music/white noise for him to fall asleep easier and to make the mornings easier when getting him ready (so that, for instance, I can use it to ascertain news/weather/daily travel info). Over the past year and a half since I put it in his room, he has of course been developing rapidly in his cognitive abilities, and he’s finally at the point now where he can wake the device on his own. It’s been several months of trying to talk to it, mostly by mimicking the simpler commands that he hears me say.
It’s interesting to me that he has grown up speaking at the Echo and slowly learning how to communicate with it in much the same way that he is learning to communicate with other people. His communicative learning progress is definitely a lot slower with the Echo than with me, but that makes sense since he spends a lot more time with people than with the Echo. Even still, I was very impressed the other day when he woke the Echo (and then promptly told it to “stop”, which has been in his vocabulary for a while now).
I’m not sure there’s any real point to this outside of just an interesting (to me) anecdote. And I guess it’s probably time I take the Echo out of his room, or at least figure out how to lock it down, so he doesn’t get into anything age-inappropriate or buy 500 cans of tomato sauce or something.
Anybody else have any interesting experiences with their little ones learning to communicate with smart devices?
Last year my daughter was just learning to sound out words and write them, and was coming to me for help in spelling them to make little sentences. One day I was busy working, so I suggested she go ask alexa how to spell the words. An hour later she came back with a whole page story! After that, she'd go to alexa a few times a week to get help with spelling, and started exploring games and asking for other info.
This is a good example of how, when guided properly, this technology can obviously be of practical and positive use for children. Sure, if you still want to criticize you could say that the child should rather learn spelling with their parents/real humans, but hey... Get used to technology already.
Imo the biggest problem when introducing children to these devices is how to get them to understand that this is very different from an actual human. Even if you find proper wording that a 4yo would understand, these words are easily overwhelmed by the fact that you can talk to it like a normal person. It's already interesting to watch children slowly grasp the concept of (video) calls, but then taking the next step and understanding there isn't a person at the other end of the Alexa dot is yet another step, because if it isn't a human, what else is it?
My kids know “Siri” as simply a computer they can talk to. They think she’s basically a robot. Mine are 6, 5, 3. Since we use HomePods everywhere they interact with them really just like they would another UI but haven’t associated Siri with a person. It’s generally some variation of playing music or turning on HomeKit stuff. But they aren’t trying to have conversations. To them is just an iPad with no screen.
I think your tale begs the question: How will your son's (or any other child's for that matter) notion of personal privacy be shaped by the use of technology from a very early age? I think we'll have to wait and see, as this will be highly interesting.
Fair question. I wonder if it might change much thought, depending on the frame of reference.
I think a lot of us grew up with siblings in the same room. Rooms didn’t lock. Parents not looking but knowing all too well what’s going on. Schools reporting on behavior by phone or by letter that as a kid you didn’t get to see.
I think for me the notion of ‘privacy’ was just being left alone with no one bothering me. It’s only as an adult that I have a clear perimeter where no one should be able to step in. In that regard having a lot of your doings leaked to parents might not be that impacting.
> In that regard having a lot of your doings leaked to parents might not be that impacting.
I disagree. It was super stressful/impactful to me as a child. Knowing my parents would be called for misbehavior at school put undue stress on me all day after getting “an orange ticket” or later, detention. I’d get punished at school then punished worse at home.
Once, in first or second grade, my uncle divulged some petty thing I said to him in passing to my parents leading to a sit down talk. I was uncomfortable and angry. This damaged my trust.
...I think a lot of us grew up with siblings in the same room. Rooms didn’t lock.
Indeed, but those were your family members. You knew them well and there was context, for good or ill.
But in this case the device is a bunch of strangers. Perhaps no one stranger listening in, but everything being processed off site and added to the profile Amazon or Google is building about you.
The device is not a bunch of strangers. There's no humans listening to "your" conversation. Though they may be listening to anonymized clips. The activity stored is also your own to control.
I think such privacy concepts are too abstract for a child to be concerned about. You'd be hurting their view of technology and enforcing an idea of Big Brother if you were to teach them a voice assistant is a bunch of strangers listening to you.
I think facebook+cabridge analytica et al, and google's collection of location data even when the user opts out (just to name two), shows that it really is a bunch of strangers listening to you.
I don't think this is a new development, privacy in voice assistants is the equivalent of your browser's search history. If anything, that made me more conscious of privacy than anything else.
Our kids know I can see anything they ask. As well as any time spent on devices. I don't see this as very different from my upbringing. And I haven't noticed any regression in trust.
It is (was?) easy to control your browser's history. It is impossible to control what these devices send to google/amazon.
Your browser's search history only gets populated when you, well, search. These devices are collecting information 24/7, regardless if you are actively engaged with them.
I don't think your last statement is accurate. While the device is "listening" for the trigger words, it's not storing information. It only collects information you call upon it for. You do need to actively engage. I have not seen evidence of otherwise.
Hasn't technology throughout the ages always reshaped notions of what people consider valuable and not valuable, with some people choosing to hold onto prior notions and others newer ones (possibly because they benefited more from such)?
To me the biggest issue is the relative one way economic (let us set aside the intangibles, because one cant feed themselves on such) proposition of these devices (upfront cost of "paying" [more like renting since its default locked in to a provider] for the device, and the free "work" people provide with their queries). Reminds me of the proposition of the collect your dna as a service companies.
I guess the incentives of everyone running an instance of sphinx, and sharing models/feedback error corrections continuously with each other in the background with the ubiquity of torrenting now and decreasing reliance on the Amazons in the middle isn't here yet.
I live with my 6 & 9 year old nephews. I set up a Google Home in their bedroom and some in the common area's. It's been very interesting to watch them use it.
The 6 year old's pronunciation is not perfect and I could see a lot of frustration when he was initially using it. The drive to use the device has driven him to take his time and really focus on pronunciation so he can play the song or video's that he likes.
They understand it's a computer. But we all tend to treat it as an assistant. A really touching moment was when he used the Home to call "Santa." After he finished that call and we walked away, I saw/listened in My Activity that he asked to call God.
It's going to be interesting to see how kids partially raised by Alexa during these critical stages of rapid development end up in contrast to their peers taught by humans.
Some obvious predictions for Alexa Babies:
1. They will develop lifelong bonds with Alexa/Amazon.
2. They will have lower IQs than their exclusively-human-taught peers and (in more extreme cases) be considered intellectually and developmentally disabled.
3. The derogatory term "Alexa Baby" will come to define a generation who's parents failed them, ultimately leading to the redefinition of "child abuse" and "bad parenting".
My son learnt how to play music on Alexa and stop it with "Alexa, stop". Unfortunately he now stops it playing music that my wife and I want to listen to - or tells me "Daddy, stop" if I am talking and he wants to say something...
This is sad. You have relegated your child's cognitive development to a cheap electronic device. He's supposed to be communicating with you and learning how to communicate with people, not crappy electronics designed to gather your info to sell you stuff.
I appreciate the criticism, but I really don't feel that my child's cognitive development has been devalued in any way due to his occasionally being able to communicate with the Echo. The vast majority of his communication is with other people - which he loves participating in. I'll concede that these devices do probably attempt to sell us stuff, but up to this point, I have not seen anything from the Echo that I would consider nefarious or that seems like it might cause harm to my child's growth. In fact, my experience is that the Echo is basically harmless, especially compared to the behavioral problems that I've seen in children who watch TV or Youtube extensively (which I do not allow).
So a child should never read a book? nor watch an educational video? Alexa and other assistants are just tools. Tools can be used against people, yes, but they also can be used to improve people's lives. This particular example doesn't mean that the parent will delegate all his parent duties to Alexa. Don't get me wrong, I think I get your point, but should I stop using knives just because they can be used in dangerous ways?
A book provides multisensory learning through sight, touch sound and smell. Alexa being voice controlled provides only auditory stimulation and is going to be the least good tool for the job. http://faculty.ucr.edu/~aseitz/pubs/Shams_Seitz08.pdf
Alexa should be treated as a different kind of stranger. I think it's a huge mistake for children to talk to computers. It's inherently dishonest for a single voice to represent something so manipulative and fluid as if it's a person.
> It's inherently dishonest for a single voice to represent something so manipulative and fluid as if it's a person.
What are you referring to when you say "something"?
It's strange that you find dishonesty in these personal assistants but then shared a subversive piece of propaganda as an "amusing" video.
The video you shared is suppose to be a friendly fireside chat by two everyday gentlemen trying to objectively discuss the manipulative subversive left leaning nature of the digital assistant. The video however is a form of propaganda and manipulation because the two gentlemen are clearly very right leaning in their ideology and sit there being quietly outraged when Alexa answers questions in a manner that is in opposition to their ideology.
To you it's a "huge mistake" for child to interact with a "manipulative" computer as if it's a person with actual opinions, and thoughts; but it's ok for you to share "amusing" videos of guys being mildly outraged that a computer is respectful of a transgender person when they clearly disagree with showing Katlin Jenner any form of respect.
If anything this video is the type of subversive manipulative bullshit that people should be fearful more so than a stupid little hockey puck that most kids find to be silly and frustrating.
It's not amusing at all. It's incredibly spot on. Alexa has very limited knowledge and 0 interpretation skills, but, worse than all, behaves as if it does.
(note that, in a sense, a book has very limited knowledge and 0 interpretation skills, but its behaviour is passive).
I was in college before I learned not to believe everything I read. Everything needs context. I tell my kids that voice interactive devices are just robots running programs written by people, and they're always at least as wrong about things as the people who wrote them. Most importantly, don't trust a robot when it tells you something that doesn't make sense. It's probably wrong.
I wish somebody had told me that about books when I was a kid.
Indeed, I hate that AI is such a buzzword when what we have are a fraction of where we need to be. We have jigsaw puzzle pieces, really tiny ones in what feels like an infinitely large (or who knows how many pieces) jigsaw puzzle.
I don't mind them being called "Voice Assistants" though, but don't tell me it's AI or ML, since it doesn't truly learn, it calibrates your voice if anything. It doesn't do a darn thing on it's own, someone has to tell it to do that thing, like in some cases the very human communicating. When they are fully autonomous... THEN I'll be happy, and creeped out at the same time.
The real change with AI will be if the masses are allowed to build their own relatively easily enough. When even kids are allowed to be creative with different AI's it will get interesting enough.
Sadly some megacorp will buy it out and lock it up is more likely to happen.
> don't tell me it's AI or ML, since it doesn't truly learn
The term AI is a minefield to talk about. But machine learning is well defined and it’s definable you machine learning even though Alexa “doesn’t learn”. Alexa uses applied models which where built using machine learning techniques. Like it or not that’s just the facts.
"Explain protocol," Nell said. This was how she always talked to the Primer.
"At the place we're going, you need to watch your manners. Don't say 'explain this' or 'explain that.'"
"Would it impose on your time unduly to provide me with a concise explanation of the term protocol?" Nell said.
Again Rita made that nervous laugh and looked at Nell with an expression that looked like poorly concealed alarm.
- "The Diamond Age" by Neal Stephenson
What's going on here is that Nell has, without even really setting out to, pierced the fiction that her second utterance is really different from the first. Humans react badly if you treat them obviously as automatons, but disguise it even slightly and somehow that's fine.
She code switches from a modality that's convenient for addressing her instance of Runcible (a powerful AI with the outward physical appearance of a large old book) to one for addressing teachers and people in authority and she does it seamlessly, so that the effect is unnerving.
Interesting. I guess that in real life both of those questions would be considered rude. The first as "too little effort and overly bossy", the second as an obvious ridicule.
I think it's a huge mistake for children to talk to computers.
I'm not so sure I'd go that far, or that it's any more dangerous than a web browser.
My kids are quite young and I'm happy for them to talk to Alexa, under supervision. They need to learn to deal with potential dangers before they encounter them without parental supervision.
Mind you, we play with fire too so maybe just ignore me.
For what it's worth, this video appears to be bull.
For most questions, it will just read you the first line (or first paragraph) of the corresponding Wikipedia page. There have been cases that Alexa has said some embarrassing things, mostly because people edit controversial Wikipedia pages is embarrassing way, but I believe they are trying to be more robust to that.
To that point, asking these same questions to my alexa:
- RE gender: same response, which is fair enough (and in line with what you'll find in most places online)
- RE Muhammed: "the founder of Islam"
- RE Jesus: "also known as (...) was a first-century Jewish preacher and religious leader.[12] He is the central figure of Christianity. Christians believe him to be (etc.)
The way how Alexa rambles (less pause during words, at full stops) the long text about Muhammed also makes me suspect this video is doctored.
But even if it isn't doctored, they definitely skipped the research step and went straight to bigotry and hyperbole...
I don't think speaking to computers is a problem so much as speaking to someone's marketing-data-aggregating cloud service. This can even be explained to a child fairly easily, e.g.:
"Siri isn't the iPhone. She's a very powerful computer who lives far away, and talks to us through the iPhone. Just like you can talk to Grandma through it."
It’s funny watching Leo Laporte’s podcasts on twit.tv. He surrounds himself with those assistants. And despite having a crystal clear, radio voice all the devices either are triggered by mistake or ask him to repeat half of the time.
Now if you are doing a tech radio show, this has a high entertainment value. If you are busy, in the middle of something and need a task done, these repeated failed interactions are a complete non starter. Imagine a smartphone where the screen wouldn’t register a tap half of the time, it would end up in a bin very quickly.
Plus I am not sold on non self-discoverable UIs. Like I don’t remember the name of all the operas I have on my phone. It takes seconds to browse with a screen. Same things for the commands. If forcing users to learn by heart dozens of commands was commercially viable, we would still be using MS DOS! But GUIs are way more powerfull.
GUI's are faster when you're near the device and don't have your hands full or when the command is complicated. But I have Echo's in various forms in most rooms now because while I also find discoverability an annoying issue there are plenty of commands that are obvious enough to learn quickly that work for the easy cases that you do all the time. A dozen or so commands covers maybe 90% of what I want it for.
Random story: my niece was visiting shortly after we got an echo, and we showed it to her and told her to ask it a question. She held something up and asked, "Alexa, what color is this?" Cue trying to explain to her that Alexa doesn't have eyes.
It's not very scientific to set up an experiment to test a hypothesis, then mess up in some way (they forgot to turn voice recognition on) and fix it by changing the hypothesis afterwards. However, I understand they couldn't let all the effort they put into it go to waste...
When you add in the idea of their constant childhood pal being a device designed to sell them items from Amazon, it's quite worrying.
If you wanted to take a positive spin on it, you could see it as the possible beginnings of a form of digital post-scarcity centrally controlled economy communism, I suppose.
I've seen my kids try this. Very weird. They want to just talk to Google assistant, rather than get anything done. Google assistant does not work very well for this use case.
"Your daughter pauses, stammers, mispronounces a few words. She’s a beginner, after all."
give her some real toys or let her play outside if she's not of the age to even speak normal. don't put little children to have some internet bot as a friend because you're lazy parent/.
Ah, come on, it’s quite witty - and it can get a bit worthy on here if someone isn’t sticking sticks into the termite mound. Plus, the commenter is using wit to validly draw attention to the ethics of habituating children to use of voice recognition tools where they are unable to consent to some of the underlying technology function. That’s a very valid wider point, but communicates it in a way which makes one laugh. More power to whoever wrote it.
Respectfully, since it's your opinion that rightfully decides, I often find it's the simple comments that cut right to the issue; that's not flamey, it's honest. Alexa could change drastically with a buyout. Would she tell us?
I think it is important to distinguish between Google, Alexa and Siri. The privacy ramifications are in bold relief. Using Google or Alexa in your home is essentially allowing privacy-invasive tracking into your every conversation while Siri is just another UI that is an extension of your devices. The privacy implications of Alexa and Google can’t be overstated. For the privacy inclined, Siri is the only option.
Not sure about Siri vs Alexa. I'm Apple fan, they are really outstanding in security and privacy, but what makes it different from Alexa? That Amazon has an online shop, thats all?
Google is a global Ad corp, with all your data, I don't even like to talk about it.
I clicked on the growingupbilingual link. About every third word is a highlighted brand. I have trouble reading it as a personal account when it looks so much like advertising.
It responds by profiling, grooming and manipulating them from a very young age.
"Aww look, my 3 year old is having a conversation with a massive advertising corporation, and has begun talking to human beings the same way he talks to an inanimate object. How cute!"
Soon, he started to apply that scheme also to talking to their parents:
"Mom, I want vanilla ice cream."
"No, you had enough."
"Mom - I - want - va - nil - la - ice - cream."