I'm not normally one to be creeped out by these kind of things. But this is a device that's designed to take photos and videos of you while you get dressed, and will be both connected to the internet and constantly communicating those images over the network.
This feels like a device literature professors would use to teach dystopian fiction writing.
what's worse: that amazon thought that enough people wouldn't care about an always-on, always-connected camera watching them get dressed (and undressed) every day, to make it a viable product ? ... or that _they might be right_ ?
On one hand the Dropcam is constantly uploading video and audio of you and storing it in the cloud for extended periods of time, even when you're home.
On the other hand, the Echo Look only uploads photos when you ask it to, and AFAIK doesn't store them, but everyone is freaking out because it could upload photos whenever it wants.
Sure, HN doesn't see the value in a Look, but it is clearly less invasive than a Dropcam.
No, because people typically either install nestcam outdoors or set it to turn it on when they are not at home. Also, I suppose very few people install Nestcams in the bedroom/bathroom/closet.
Then I'd bet you're not in the target market for a camera that can record you when you're trying on new outfits.
But if you were, I'd be willing to bet that you have a laptop in your bedroom or phone-cradle that can hold your phone in a position to record you so you can instagram today's outfit to your friends.
Even if you're "in the market" for this (and not everyone filmed will necessarily be a buyer - e.g. spouses, kids, etc), it's still riskier to have a camera always pointed at you when changing clothes than one you usually have to purposefully point at you.
(1) This device is absolutely sealed and opaque. It can't (easily) be analyzed to see what it's running, doing, or sending. It's even more opaque than a phone. It has enough processing power to do local image analysis, object recognition, etc., so just looking at bulk traffic won't prove anything about whether or not it's watching.
(2) The camera in this device is part of its core functionality, is always on, and is situated deliberately so as to watch you.
(3) The device is literally an Amazon sales rep in your house. Its primary purpose is to represent Amazon and sell you stuff.
I've always thought Alexa and the other cloud-connected voice-activated things were completely insane. I will never own such a thing. Period. This takes it to a whole new level of creepiness.
Well, not completely opaque -- it uses your wifi network, so you can watch its traffic to see if it's sending streaming video back home without your permission. The stream should be encrypted, so you may not know exactly what it is sending back home, but you simple traffic analysis will tell you if it's sending video.
Having it being completely sealed without the ability of the end-user to install software is a security win for most people. Then it's much less likely that a user will install Malware on the device, and it's not likely that Amazon is going to hack their own devices to make them into spy cams. Not allowing user installed software greatly reduces the attack surface of the device.
Even if it wasn't intentional, unknown bugs are a thing, and trusting someone like amazon, who clearly does not have their customers best interests in mind, seems foolish to me.
Amazon says their camera only records you when you ask it to, if you don't trust them, then how do you trust that your cellphone or laptop isn't recording you illicitly?
How do I know my smart phone isn't shooting a video of me when I'm dressing? It isn't pointed at me. It doesn't sit on a stand with a view of the room. Sure the microphone might catch something interesting in the bedroom, but setting up an internet connected camera with a view is very different than than one that is pointed in a safe direction.
Firearms: never point it at something you don't want to destroy.
Internet connected cameras: never point them at something you don't want the world to see.
I don't bring my phone with me when I'm going to shower or change clothes. If I do, I can place it face down and even place something on top of the rear camera if I'm afraid of capturing something other than my ceiling.
This Amazon device either needs to be left alone to capture/not capture whatever it is going to or it serves no purpose at all. I'm not going to turn it around or cover it every time I enter my bedroom. If I did, then the value proposition falls significantly.
It's not even an issue of trusting Amazon. Or, just trusting their motives. You also have to trust their security on an ongoing basis, as your device could theoretically be compromised at any time.
While this may be true, I'd trust Amazon more than the various Chinese camera makers that offer cloud connected streaming and storage. Heck, I'd trust them more than my own ability to secure a vulnerability-laden home router that AT&T manages and probably never patches.
The point is that one doesn't usually purposefully keep our cameras directly pointed at us when we're changing clothes. Cellphones in particular are usually pointed at the table directly under them. Laptops are often more risky, which is why these exist: https://supporters.eff.org/shop/laptop-camera-cover-set
>>if you don't trust them, then how do you trust that your cellphone or laptop isn't recording you illicitly?
I dont, and I dont trust them, I am one of "those people" that places tape over the lens. I also do not own an echo, nor will I own this or any device like it
I hate that Phone do not have removal batteries because there was a time where I did remove them.
Some may call be paranoid, but only the paranoid are ever truly prepared
The products aren't exactly comparable in terms of utility. I get a whole lot more utility out of a cell phone or laptop then I would out of the amazon look. Even if both have security risks one has substantially greater rewards.
That's true. I was more saying that it's not usually pointed in my direction when I'm changing. I have a cover over my webcam and my phone is usually looking at the ceiling or whatever surface it's on.
My wife does this daily thing where she tries a pair of clothing, looks at herself in the mirror, sometimes ask my opinion, then change into complete different outfit, repeating the same process until she settles on what she would wear.
I swear, my wife would LOVE this product, unless I can scare her about the security implications.
There are Kinect-based 3rd Party fitting room applications, mostly used as demos at fashion shows it looks like from a quick search. There are also Kinect games that play with similar concepts, in the space of a game.
The Xbox itself has never had fashion/fitting advice as a feature.
> this is a device that's designed to take photos and videos of you while you get dressed, and will be both connected to the internet and constantly communicating those images over the network.
The HN reaction to this device is what I expected: we're somewhere between suspicious and terrified.
I'll be really fascinated to know how the wider market reacts. I was one of those who believed consumers would look at the always-on living room microphone with suspicion. But I think the lesson was pretty clear that people will trade privacy for convenience. (Likely an overabundance of trust or a lack of understanding of all the real implications and potential for exploit.)
One bit of perspective: When I was a kid, a photo was kind of big deal. My kids take and share photos of themselves every single day. I'm not sure they consider this to be the same privacy invasion that we do.
"(Likely an overabundance of trust or a lack of understanding of all the real implications and potential for exploit.)"
I think that's a lot of it. People aren't trading privacy for convenience... people are taking convenience. It makes a lot more sense when you think of it that way.
You can do an experiment yourself to show this; take 5 non-techie/non-HN people and explain to them exactly what this thing does and how it works. You'll find people's attitudes change, rationally or otherwise, which I present only as evidence that they haven't thought about this and incorporated it into their worldview, not that they are right or wrong. Or watch this: https://youtu.be/XEVlyP4_11M?t=24m54s John Oliver on Government Surveillance.
Perhaps ironically, mere nude photos of myself aren't really what I'm worried about. I mean, I wouldn't be happy if they got out, and since I'm not in the habit of texting them about I have questions about where they came from, but by and large, nobody would care and it would affect my life only briefly. I'm far more worried about the government scraping everybody's postings on Facebook and building up a GoodThink/BadThink database, which I know basically already exists, because our political parties (let alone our intelligence community) build voter databases that are probably at least 90% accurate on that front. A concrete list of Everyone Who Disagrees With Me, one accurate enough for someone powerful to decide the rest of the inaccuracy is acceptable collateral damage, is scary.
Yep. I'm going to London on Friday. The Mayor of London just decided that anybody who posts stuff against muslims on social networks is guilty of a hate crime. I am thinking I won't get arrested when I arrive... but it's a possibility, and I definitely won't access twitter while I'm there.
The leaks coming out from Wikileaks over the last couple of months has been amazing. All this stuff that security people have theorized about is not only real, but in wide application, often with vendor complacence if not out and out cooperation.
Do you have a Samsung TV built in the last 5 years or so? The USG has turned _that_ into a permanent listening device, even in "off" mode.
Much better target than the Echo, which people already know is a microphone that streams out a live feed of audio from their room, and may unplug if they're feeling paranoid.
I wonder if something like Sense [0] could help ensure that an "off" device is really off.
Totally off topic: Yeah so, about that 90% number... You you really think that is the best they can do?
I mean, 90% is pretty good for, like, dog food ads and car commercials, but with lives, a 90% true-positive rate is just garbage. Like, with all that terrific amount of data, are they still just doing T-tests, ANOVA, MW-U-tests? Like, what is their p-value, still 0.05? I know this is super stats-wonky for this thread, but I mean, come on, they have to have some super secret stats and mathy stuff that they are doing, right? Like, formulas and theories that are just really good. It's been, like, 15 years they have had this scale of data, and it's only growing, right? If so, nothing at all has been sent out to the academic community, which, for math theorems, is kinda hard to believe. I now signal-to-noise is super important for NatSec, but it's also super important for DrugDev.
But yeah, a camera that is meant to watch me dress and then order shit for me, that is a super no-no. It, like, actually gives me goosebumps.
"Totally off topic: Yeah so, about that 90% number... You you really think that is the best they can do?"
For the political parties specifically, yes, by the standard they care about. Remember how we're always talking about how the centrists generally end up with the deciding vote, and how the polls are oscillating around by ~10% in the several weeks leading up to the Presidential election? Those people themselves don't really know who they're voting for or whether they support the "correct" person, for any given definition of support, so it's a bit much to expect anyone else to accurately guess.
In the event of a true police state which gives you the choice of vigorous fealty or the Gulag I would expect they can go much higher and that the initial competent execution would be almost inescapable. (Over time it would decay due to various forces, but I can't put a very solid time frame on it... between 10 to 50 years to develop very serious holes in it, probably, but even 10 years is an awful long time.)
I've read that they won't be using them, because they generally have professional setups with proper lighting and DSLRs. They also then edit photos before posting them.
I put it in quotes because it's become a kind of catch phrase, and in practice I have found this to be largely correct. We've tried very hard to make our product secure, but users almost never ask about that. We have never been asked about the quality of our crypto by a regular user, and only once by an enterprise user. Users care about UX and cost, in roughly that order, and little else. This is even true of most enterprise users.
This is why security sucks. This is why everything spies on you. It's not a priority for anyone because it doesn't affect user buying behavior. The tiny minority who do care are not only too small to matter but also tend to be the kinds of "hackers" who like to DIY (and pirate) rather than buy things... making them doubly economically insignificant.
In the end the privacy issues around things like this will have to be fixed with legislation. We will need to effectively extend HIPAA regulations to cover all kinds of other personal data: passively recorded audio and video, location tracking data, anything more than the most superficial user logging, etc. Vendors will face a fine of $$$ per incident if this information is leaked, and sale and use of information will be strictly regulated.
As far as government surveillance goes: that horse left the barn over a decade ago, and that also can only be fixed in the legislature. The legislature must regulate intelligence agencies and police. If they don't, no amount of techno-fixing will limit the power of these agencies. They have larger budgets than you.
> I was one of those who believed consumers would look at the always-on living room microphone with suspicion. But I think the lesson was pretty clear that people will trade privacy for convenience.
I don't think it's a matter of trading privacy for convenience, as long as you believe Amazon when they say they're not sending anything when you're not using the wake word.
All evidence points towards this being the case, both my own anecdotal evidence, and the investigations people have done.
Can it be exploited? I don't know. I'm not going to speculate.
I don't see any additional trade off vs. having my cell phone on me at basically all times.
Yes. People don't see this as a device that's "always listening". They see it as a device that has a voice-activated "On" switch. Instead of pressing a button, you say a word.
Of course, with a little bit of thought, you can connect the dots and figure out that it has to be listening all the time to know if the wake word is said, but most people aren't thinking that deeply about it. When they do realize this, Amazon has assured them that the "voice switch" actually does function as advertised and that audio is not streamed until the device is "on".
This is quite straight forward, and I'm surprised no one has done it (to my knowledge).
You simply wire the camera in such a way that powering it also powers the LED. If done properly, you couldn't hack this with software.
Google could hide proof of either statement by manipulating its search results.
I'm not suggesting this is actually happening. I'm pointing out that when you have a single distribution channel for information, users have no choice but to assume the channel hasn't been tampered with.
Of course you could have papers, conferences, and so on proving otherwise. But if they existed, you would only hear of them by word of mouth - which would rely on being in or around a professional network which took an interest in such things.
If Google decided to censor or manipulate the search results about a topic, it could become extremely hard to access reliable uncensored information.
This is a brilliant product, and right in the corner of a "brilliant product/things hackernews hates" graph.
Most people aren't that concerned by privacy and will love features like the ability to see different views and short videos. Everyone knows that feeling of catching a glance of themselves in a passing mirror and how it's a different feeling to deliberately evaluating your look in a mirror. If they can recreate that feeling, and combine it with product recommendations it will be a massive success if they market it right.
> When I was a kid, a photo was kind of big deal. My kids take and share photos of themselves every single day.
It makes me wonder - is it no longer narcissistic for people born in an age where camera sensors are ubiquitous to take selfies constantly? Can a behaviour be normalised as a result of cost, the same way listening to music in public would have been considered indulgent and neglectful of the world around you before everyone could afford a personal/portable music player?
Personally I lean towards it still being a narcissistic trait. That's not to say the behaviour isn't natural - it's no more artificial than our tendency to gorge ourselves on junk-food because of its cost and availability. But it's something that needs to be controlled.
> It makes me wonder - is it no longer narcissistic for people born in an age where camera sensors are ubiquitous to take selfies constantly? [...] Personally I lean towards it still being a narcissistic trait.
Well, people used to get 8 foot tall oil portraits of themselves done; a few extra digital photos still seems a pretty big step in the right direction.
I suspect that if commissioning an artist to make portraits of any size was as cheap and easy as taking a selfie or eating junk food, we'd be balls-deep in portraits right now. Cost & effort vs value.
The implication being that narcissism is nothing new or changing in prevalence, we just need to discourage its more overt forms so that we have a healthy and thoughtful population.
I think we've done a pretty good job about this, what we're left with is people taking pictures of themselves costing $0, and perhaps posting them on free sites for people who want to view & share such things. That's just about as out-of-your-face as you can get.
Overt meaning publishing our narcissism uncensored, rather than acknowledging that we're naturally concerned with our outward appearance and status but see the value in balancing it with some modesty lest life become a silopsist shitshow.
You could argue that a painting, while expensive, is less overt in a spatial sense than making images of yourself available to practically any person in any location at any time. It's multi-dimensional vanity repeated ad nauseam because it costs practically nothing. Paintings are limited to specific times and locations in space-time.
It's completely not for us ("us" here being people not super into fashion, since this isn't really a fashion website), I wouldn't think, which is why we look at it cockeyed.
Yes, I have a teenager at home that would probably enjoy sending pictures of outfits to friends to check out. There has even been talk of a outfit calendar so that the same one doesn't get worn in the same couple of days|weeks|months.
I would have absolutely zero use for such a thing, but HN is a pretty small community and there are a lot of people in the world.
> But I think the lesson was pretty clear that people will trade privacy for convenience. (Likely an overabundance of trust or a lack of understanding of all the real implications and potential for exploit.)
Or perhaps more accurately, most of the people in question are members of the mainstream dominant culture and don't really have much if anything in their life that they worry about someone else finding out about.
They're not, actually. Almost no one is average if you look at multiple features at once. I would guess that most people have something they wouldn't be too happy with everyone knowing about.
"In his research measuring thousands of airmen on a set of ten critical physical dimensions, Daniels realized that none of the pilots he measured was average on all ten dimensions. Not a single one. When he looked at just three dimensions, less than five percent were average. Daniels realized that by designing something for an average pilot, it was literally designed to fit nobody." [0]
This is meaningless without considering the granularity of the unit of measure. Most people are of average height when measured to the nearest meter. The motivation behind the measurement determines the granularity -- for a cockpit, the seat and pedals and stick probably have to be within a couple of inches of tolerance. For a society, how racist or treacherous or murderous does a person have to be before we no longer accept them in society?
I think there are simple guidelines: first, make clear the distinction between public and private spaces. No mandatory or secret surveillance in private places. Second, informed consent and transparency: Those willing to give up privacy for convenience within private spaces must have access to (but not be forced to read) clear, understandable, honest descriptions of what their information devices are doing.
But we don't treat people by average very well. We treat them bad if they have any kind of psychological issues. We treat them bad if they have weird sexual fetishes. We treat them bad if they have the wrong kind of friends.
Just because it's "average" doesn't mean it's safe to be public knowledge
I get your point, but has the "general consumer" actually demonstrated that they will trade privacy for convenience with regard to Amazon Echo (and competitors)? This stuff is still very early days. I don't know anyone that has an Echo, even among my tech literate friends. I think it is yet to be seen if this type of technology really gets universal adoption.
I know at least a dozen non-techie types who have an Echo or Echo Dot in their kitchen. Mostly, they use it to listen to music, help with timers/conversions while cooking, and to hear the weather/news in the morning. Not one of them even thought of the privacy concerns until I brought it up.
It's consensual or whatever, but I think the bigger factor is that people are not really aware of how much information can be gleaned from the photos and don't consider the implications of Amazon storing them forever.
Even trying to see this from an outsider perspective it seems like a stupid product. As someone else pointed out, they've replaced looking in the dang mirror with a multi-step process involving a few hundred dollars of tech at least. They made sure to throw in "cloud" and "machine learning", but the average consumer doesn't care about that.
Amazon has had great success with some of their stupid products in recent years. Maybe for you and I We just want to throw on some clothes and get out the door, but for some getting dressed is a ritual. They don't want less steps, they want more. I've had GFs who'll spend hours trying on different outfits.
And the next logical step would be a virtual dressing room, that uses an augmented reality overlay of the sort we see with Snapchat's funny faces toys for trying on virtual clothes, which would convince people to buy more shit from Amazon.
> And the next logical step would be a virtual dressing room
Hey, that was my old startup idea!
Also, I kind of wonder why this still isn't a thing; the technology for doing that cheaply has been around for at least 7 years[0], and just being able to quickly view many different outfits on yourself would be useful enough that even I (a totally fashion-oblivious guy) would consider using it.
Still, in a perfect world, such a thing would be implemented entirely locally (again, the tech required for doing that is old), with only clothes database being downloaded. This is exactly how this won't be implemented - it will instead send data to cloud for various anti-consumer, business-friendly reasons.
> And the next logical step would be a virtual dressing room
Another logical step might be skin-cancer screening (surprised that, in 300-ish comments, no one seems to have thought of this), assuming the camera's resolution and/or back-end image-processing software could be made good enough to pick out moles, etc. After all, the camera could take full-body pictures of front, back and sides very easily. Then, if anything odd shows up, you could perhaps ask the user to do a close-up shot of the affected area for further examination and/or just be told to go see a dermatologist.
While fully-naked shots would be best for this -- with all the attendant privacy/HIPPA/etc. issues -- I'd expect even doing this while in underwear/bathing suit/etc. would still be a net-win since you'd be able to examine around 90% of the body and you could do it regularly. How often, after all, do people normally get any kind of dermatological exam?
(Mind you, a dedicated device for skin-scanning might be best, but, as they say, the best camera is the one you have with you -- if this sort of device becomes popular, then it might be good to make it work for this vs. letting the perfect be the enemy of the good, as they say.)
I think you may be missing the brilliance of this product from a sales perspective.
Even if this product isn't a massive success, the units they do move will have a huge potential impact to their retail clothing lines.
Now, instead of suggesting clothing items in the store based on other items you've looked at, they can now suggest clothing items in their sales material based on what you already own and look good wearing.
If I had this product and Amazon started sending me curated outfits that they knew would look great on me and matched my style, I'd probably start using Amazon exclusively to fill out my wardrobe.
As it stands now, I buy almost everything EXCEPT clothes through Amazon and prefer to let my wife buy me things she wants me to wear.
This product, on it's own, is a game changer in the fashion industry. Now Amazon has a compelling and unique product that is going to make it increasingly difficult for other retail outlets to compete in the same space.
And once we start seeing "apps" for this camera that auto-instagram/snapchat your pictures, I think we're going to see a huge demand in the teen/millennial female market for what is kind of a geeky niche product.
Put this way, I'm interested. If it could reliably figure out my size and show me clothes that it guarantees will look good on my body type - I'd probably be interested. Although I'd want to just rent it, not own it.
I'm not sure what's creepier - the privacy issues, or the fact that everyone seems to be assuming that machines can make better fashion choices than people.
Is it that hard to imagine that a machine can do a better job than at least some people, particularly if the machine is trained by experts and/or Big Data and/or even Mechanical Turk? There's a reason we have terms like "fashion-challenged" after all.
Also, even if its judgement were about as good as a person's, it might be helpful in being able to quickly render that judgement vs. people to-ing-and-fro-ing. A machine might be able to even provide explanations for why it thinks something looks better than others, e.g. "While this outfit fits you well, it does not look as good on you now as it did 5 months ago -- its light colors make you look very pale, possibly because it is currently winter and you are not as tanned as you were then (or perhaps you just have a cold) -- better to save it for summer months".
I was going to spend a little while spitballing about how it shouldn't be all that hard to build a wall-hangable replica without all the creepy cloud-bound spy smarts - and then I found this [1], which does a better job than I could. I'd probably clean up the sides and use something nicer than veroboard for the grille, if I were making this myself, but it's a very good first approximation to say the least.
And Adafruit's method solves the major problem, of where to find a suitable lens to replace the rather expensive [2] original prop part, very handily! Unfortunately, they're out of stock of the rather crucial button, but Sparkfun appears [3] to have no trouble sourcing them.
ETA: On further reading, I wouldn't follow this Adafruit method; I don't have access to a laser cutter, and even if I did, having the tabs stick out the side, and the whole thing sort of jigsawed together that way, doesn't appeal.
Instead, I'd work up a frame from aluminum bar stock, which isn't all that much harder to work with than plastic, and build the faceplate to mount picture-frame-style in a groove milled into the inside face of the bar. If you've got access to a laser cutter, you probably also have, or can easily enough get access to, a milling machine, whether CNC or manual - they even make mini-XY tables with Dremel mounts, which might actually be preferable for a small job like this to something more like a Bridgeport machine. (If you do want some smarts in there, you can have as much room for them as you need - just pick a suitably sized bar stock. The groove will be near the front edge in any case, since the prop doesn't have a lot of depth there; the rest is just trading off between how much space you have behind the faceplate, and how proud of the wall you want the finished item to be.)
Also, since I (again) don't have access to a laser cutter, I'd cut down the grille from the door RF shield out of a scrap-heap microwave. If you don't have one of those lying around - and why would you? - your local junkyard does, and who doesn't love a trip to the junkyard? Shouldn't cost more than a few bucks; if you bring your dikes and don't mind maybe having to stitch up a hole in a pocket, you can probably cut the piece you need and smuggle it out without paying a cent, although you probably want to be paying for something else at the same time so it doesn't look too sketchy.
Looks like a fun project, in any case, laser cutter or no!
I used to run the HAL Project screensaver, pity it was flash-dependent. It's all fun and games until your personal digital assistant decides to murder you in hibernation.
I wouldn't buy this, but thinking of how many friends seem to have a near-constant Snapchat stream of "what do you think about this hat?" photos makes me think this is going to sell big.
It's not more vain than going after the last JS framework or busting your ass in the latest SF startup that's totally going to be a unicorn.
You're not the target, and that's ok, but there's no need to be condescending to the one that might be interested
You might be on to something. But the video on the website is really notable to me for using real people (not abnormally attractive women). The clothes also look fairly normal. They seem to be going for some kind of genuineness here.
I think there are some pretty compelling arguments for this device selling like hotcakes in this thread, and my experience with people very into clothes (e.g. not myself) would lead me to believe there are a ton of people that'd love any reason to make their morning/evening routines longer, and even more people that couldn't care less about a camera/internet-enabled device set up in the bedroom.
I imagine it will take off a reasonable amount at $200, and then really start selling when it comes down to $100, a la the standard Echo path.
Having the button to refresh my cat food/litter every month was absolutely awesome. I already bought the same kind every month on Amazon (they didn't have a subscription offer) so it was always awkward to remember to get back to a computer later to get on Amazon to buy more cat food when I noticed I was low -- being able to push a button to order more right then is surprisingly convenient.
Same here. On the flip side, there is a bra-fitting app called "ThirdLove"[1] are doing this kind of computer vision/machine learning AI to help women for years. So Amazon Echo Look is not completely out of the line here.
> Style Check keeps your look on point using advanced machine learning algorithms and advice from fashion specialists. Submit two photos for a second opinion on which outfit looks best on you based on fit, color, styling, and current trends.
Taking narcissism, insecurity, and invasion of privacy to the next level. Well done, Amazon!
Now that we're replacing mirrors with cameras, what's next? Covering our windows with screens that show us what's outside? Maybe they can use machine learning to judge the weather and recommend outfits based on current trends and conditions!
The Juicero inventors are going to be kicking themselves when they see this thing. "Replacing a simple everyday process with an over-complicated piece of technology. Brilliant! I can't believe we didn't think of that!"
I mean, i love all these tools, i just don't want them to have any form of internet access.
It's why i'm building my home cloud and automation system, but only doing so if the devices don't leave my firewall. I enjoy the tech, i enjoy the complication (to a degree), it's fun gadget stuff, it's nifty. However i do not trust them to keep my information safe, even if i trusted them to not misuse it themselves (which i don't).
>> I mean, i love all these tools, i just don't want them to have any form of internet access.
I've come to the realisation that the only way I will have a 'home of the future' is if I build it myself and keep it offline. I love lots of the IOT products and when they eventually come down in price they would be no brainers. But my privacy isn't worth the benefit they provide.
Outlaw the kind of data collection and abuse that keeps these services alive and plenty of reasonably-priced paid encrypted-cloud and/or locally-hosted options for all kinds of these free spying-enabled services would pop up quickly. The spy-vertising economy, which lives mostly on people not realizing just how bad it is and coordination problems in trying to counter it, makes it impractical to compete using other business models.
$199 device that sends data to some servers elsewhere and gets the job done in seconds
$2000 device that takes up space and electricity to do the same thing, but in minutes
I personally like the second option a lot better, I wish we had an open source version of all the google apps that did it's own ML on your own servers.
Except most of the IoT devices on market would do the same thing for about same price without the cloud. The only problem the clouds solves for users is not having to know how to set up a VPN so that their shitty app can view mostly meaningless pseudo-charts when the user is away from home. But in exchange for that, you give the vendors lots of data to resell, and the ability to brick your devices remotely.
SaaS model being ported to hardware is probably the single most user-hostile development I've seen in recent years in tech.
There's a much more immediately practical reason to keep it in-firewall as well: if the utility company accidentally cuts a fiber line outside my house, I don't want to be suddenly unable to turn off my hallway light.
Wait, so now building codes are going to get involved with what I run in my house after it is built and inspected? That doesn't sound like building codes. If anything, regulation from the FCC or someone else would be more uh, fitting.
Sure, I was thinking of things like lights supporting local control. I guess I'd try to write it so that a phone app talking to the light counted as passing the rule, as long as it worked without an internet connection.
I don't see how it is a problem for simple rules like "Installed lights must always be controllable from inside the premises" to be code.
Caring that your clothes look good is narcissism (excessive interest in your appearance) now? Have you ever asked a friend if they like the suit you just bought? Do you go to work in a burlap sack?
Knowingly and willingly sending people photos of yourself is an "invasion" of privacy?
> Caring that your clothes look good is narcissism (excessive interest in your appearance) now?
As if caring is a binary thing and not a matter of degree? When you pay $200 to have "fashion experts" (or AI) judge your outfits on a daily basis, have you not crossed a line?
> Knowingly and willingly sending people photos of yourself is an "invasion" of privacy?
Yes, I'm sure this won't be used to collect all kinds of personal data for marketing purposes.
I would pay well more than $200 to hire someone knowledgeable about fashion to design and help me buy a wardrobe. I don't think it's narcissistic at all to want to look nice.
You've never once asked one of your friends or spouse their to help you pick between two shirts?
I mean, for a human-powered service that just consulted me on good looks according to detailed specifications (I under-dress for work because it reads like an asshole power move, House MD style, that has served me well at least a couple times) and in accordance to my wife.
Maybe they'll topple Facebook's leader-status on narcissism, insecurity, and invasion of privacy? This definitely falls into the category of tech designed to make you sad/insecure (in this case, so you'll buy more clothes from Amazon).
I can't help but feel that we're increasingly in one of those movies where the real world becomes mixed up with a fictional one. If you were to take some of the things I've read about over the past month or so and challenged people to tell you whether they were real or a parody they'd be hard-pressed to decide.
I don't think I've ever been less the target audience for an Amazon product. But I bet this will sell a lot.
I grew up in a world where you would never even consider taking a picture of yourself to show off how good you looked to your friends. Even now, living in a world of selfie-sticks designed to do exactly that, with a population full of people doing exactly that, the idea makes me cringe inside and imagine what would happen if Jason's High School from the 80's discovered a video of Jason from the 80's using this device. It would have "ruined my life".
Interesting to watch the world change. If a bit horrifying.
That's what I'm curious about though. They went through the pain of designing a whole new hardware with a camera on it, and put a nice 200$ price tag on it, but right now, the only use of it is to take pictures of your outfit...
Paranoia about privacy aside, there are so much more I could think that an assistant could do if it had eyes, and I'm sure they may very well add those in the future, but right now, even if it was a fantastic product, I still don't see how big of a usebase it can get. Seems very niche.
Amazon Says: "We're going to blur the backgrounds so your outfits pop"
Amazon Does: "We're going to identify everything in our field of view, and filter this into the profile we've built about your possessions and interests to enhance our recommendations engine.
I'd say because putting some of our brightest minds, and best technology, to work to even further push demand for things people don't need is doing the exact opposite to what humanity needs.
Marketing has long been about manipulating people to create a perceived need, but this is stepping it up to new levels.
If capitalism wants to survive it's going to need to find a way to monetise things that are good for humanity and Earth, that's not this.
Pushing people to depend on fashion for self-worth, then dictating rapidly changing fashions in order to sell more goods ... What's good about it?
Because there's the massive potential for some of the data collected in these photos to reveal things about people they don't publicly share with friends, family or companies. The customer thought they were just buying something to get some advice on what shirt to pair with their pants. Instead, they've unwittingly exposed themselves to Amazon. If this is Amazon's intent -- and I suspect it is -- they should just be up-front about it. If it is, I'm sure it's buried somewhere in thier TOS. Make it opt-in.
Giving ammo to companies spending billions of dollars to subject you to non-stop psychological warfare aimed at affecting your spending habits (and any other emotions/behaviors it needs to bulldoze over to get to that goal) is just fine, then?
This isn't the place for 4chan/reddit style low quality content like this.
Reducing someone's legitimate complaint about using a networked device designed to give fashion recommendations to profile their behaviors and categorize/log all their possessions for analytics and marketing to 'you're just afraid of the future' is ridiculous. In any other industry this would be a flagrant ethics violation.
ITT fashion illiterate techies doubting the success of a pretty genius product
(Almost) nobody cares about privacy. But (almost) everyone cares about how nice they look in public. And their Instagram. They are not like what I'm coming to realize the average HN user looks like or behaves. Use your imagination freely here...
The main problem I see with this is the quality of recommendations. Amazon only carries certain brands/looks/styles; will it try to suggest shitty fashion? Both the male and female models used on the product page were dressed awfully. Will Amazon just try to go for the lowest common denominator here?
The best way I see this working is for the device to categorize users into certain style categories and then recommend clothes out of that category. But I'm wondering if the ML is smart enough to recommend CDG pants and MMM knitwear for one user and then A&F for another.
Remember, your algorithm's quality is only as good as your training data and if they're training on outfits like they used on the product page...
> ITT fashion illiterate techies doubting the success of a pretty genius product
I love that caring how you look is just "narcissism" now. :)
The other problem I see w/ the ML is that, due to the nature of data sets, it's going to be mostly white people who benefit. Minorities famously do poorly.
A great many thinkpieces will be written about this, but let me beat someone to the punch:
Foucault's Panopticism is the obvious tie-in. Usually when we talk about the Panopticon in the modern era, we're talking about mass surveillance or the performative act of social media. When we act in a certain (digital) way, posting this link, for example, we signal our values to the community. "You shall be known by your works". However, this kind of technology makes for an even more interesting and more troubling application of the Panopticon.
Foucault says that the group eventually becomes its own control system without the need for the central observers, that we are all at the center of our own panopticons. We watch others unfailingly, assessing their actions and putting them back in line when they stray. As we are socialized, we become more entrenched in our system of values: wear makeup, don't get fat, don't smell like cigs, don't wear stripes and plaid, and for god's sake: don't look like you're trying to hard to conform to all of this. We are both observer and observed, traversing endless nets of social expectations.
Now, instead of your peer group enforcing social expectations, we have got machines in on it! The result is something blatantly dystopian - a system that will correct you in your home before you dare venture out looking like that and embarrass yourself in front of the others. As it learns, it will learn even more that we value large breasts (but also modesty), pale skin, sleek clothes on slim waists, bodies with bulbous lips and thighs and pastel face paints. It will advise those without a perfect body to get one, and it sure as hell will provide the means by which to do so: consumption. New makeup, new tights, new corsets that you can clamp so tight it'll squeeze you into a form the camera will approve, new ways to take selfies every day to show your followers that you care enough to buy this piece of surveillance equipment so that you can "look good" without those pricey fashion magazines and stupid blogs!
there is a place to argue that when machines enforce social expectations over fleshy peers the panopticon analogy no longer applies, or has to move to other domains. When there is not an invisible human or culture on the other end of your panopticonic cell, the peering machine eyes that give you fashion/social/love/life advice have the same affect on the newly adapted society as the walls themselves had on the society before it. And the level of harm (implied by "dystopian" label) is debatable. I would argue that the human condition and its cultures will continue to exist largerly as usual so long as we are able retain 1 secret. With each new technological shift the unadapted generation thinks the last secret has been lost and dystopia has finally arrived. So its worth taking predictions from techo doom with a grain of salt.
Regardless, this product is absurd, niche, insecure and as a result will probably have little sway. Amazon would literlly have to give it away for free and subsidise the cost on some uber (pun intended) hope of monopolising and destroying the entire brick-and-mortor fashion store industry to have any hope of significant impact, imho.
There's a lot of negativity here. And maybe all of you are right. But it would be interesting to think of the most positive , imaginary scenario:
1. Amazon builds the service such that brands loose their power. All the negative psych/social phenomena of brand culture are gone, and with it, btw , all the ads.
2. High security + physical on and off switch make for an improved privacy situation , especially if you consider the current state of webcams.
3. Amazon builds the style consulting service as a positive service, consulting with feminists and psychologists, not using dark patterns, but to the opposite, using empowerment regarding bodies, etc.
This may also building a fashion social network focused on positive feeling, maybe by limiting to the close social circle, or something else, because the goal here isn't attention, like facebook.
4. That service would be good, it would really help women look their best.
5. Clothes become more affordable.
Now, if Amazon could do all of this, it would be really great for their business, so why wouldn't they try ?
> 1. Amazon builds the service such that brands loose their power. All the negative psych/social phenomena of brand culture are gone, and with it, btw , all the ads.
This is extremely naive. Amazon's goal is to sell you more stuff, including clothes. In the long run this device will regularly tell you that you need new clothes because your outfits are no longer fashionable. Even if that's not a feature now and even if this version was created with the best intentions someone in marketing will see the power of this thing to sell people more stuff. It's their job.
Edit:
> Now, if Amazon could do all of this, it would be really great for their business, so why wouldn't they try ?
How would what you listed be great for their business? Selling more stuff is great for their business and the best way to get there is to tell you that the stuff you already have is not good enough.
The reason why Amazon and more people don't do these things is because they are hard and success breeds expectations that can only hurt you.
It's much cheaper and safer to build a politically neutral brand without expectations of some kind feminist message / worldview. If you go that route, it isn't going to sell more clothes but, it will invite people's hyper-sensitivity about every aspect of the product. You can only lose in that scenario, you aren't winning anything.
Also, it's creepy that we replacing a task that friends used to do, (fashion / styling / taste advice) with a corporate cloud algorithm? Seriously, just get some friends that like fashion and ask them to rate your style. I guess it's a bit needy to ask them every morning but, then I suspect that most people don't have that many clothes and that many possible looks. Also, looking good is zero sum game. Ultimately, you just competing for the same pool of attention from your friends. If you are successful and soak it all up with your great Amazon(TM) fashion advice, then your friends will soon have it too and then you are back to where you started but, now you have this dependance on the kind of friend who gives you great advice (that you need to just look decent now, not better than average like before) except it keeps encouraging you to buy more fashionable (and thus more expensive) clothes. It's kinda a shitty friend at that point.
As for the question if political branding, maybe Amazon doesn't brand it this way, just says "data indicates our customers will love this service"? And customers ,coming with zero political expectations get all this benefit and are happy ?
Now that you've done the optimism, let me help with the pessimism.
>1. Amazon builds the service such that brands loose their power. All the negative psych/social phenomena of brand culture are gone, and with it, btw , all the ads.
Brand culture is alive due to marketing / advertising. $AMZN wants to be paid to market / advertise. It's more likely the case that brands will be able to pay to have their brand identified via the cam (giving additional points) or have their brand (in a similar style) recommended to the user via the app / website.
>2. High security + physical on and off switch make for an improved privacy situation , especially if you consider the current state of webcams.
Agree, but a physical on and off switch makes the biggest selling feature useless: speech. Alexa loses it's cool if I have to constantly turn it on and off. And, heck, at that point, why not use a normal camera?
>3. Amazon builds the style consulting service as a positive service, consulting with feminists and psychologists, not using dark patterns, but to the opposite, using empowerment regarding bodies, etc. This may also building a fashion social network focused on positive feeling, maybe by limiting to the close social circle, or something else, because the goal here isn't attention, like facebook.
You will not convince me that this point is not in jest. The product is not a "feel good at any size" product. If it were, it's have a small utility for an even smaller group - this product is to _increase_ the importance of "fashion".
>4. That service would be good, it would really help women look their best.
I don't see how this is the conclusion from #3.
>5. Clothes become more affordable.
It wouldn't do much. $AMZN doesn't have much power in the fashion market.
It's for placement. You can't hide it out of the way or point it up at just your head. They need a good look at your whole room and your whole body, right down to the shoes.
I'm 95% confident that I have no use for it and expect complete failure, therefore it will sell and cement Amazon as the #1 player in machine learning.
I was fairly skeptical as well, but then I showed this to my friend and she thought it was a great product idea (albeit potentially overpriced). Apparently a lot of people like to take fit pictures and also log outfit combinations which work for them.
It's absolutely brilliant if your goal is any of several things that don't actually involve delivering a valuable service to the people buying the device (if you must incidentally, that's fine).
And if you have never so much as encountered the word "ethics".
It doesn't need ML, it will always be ever so slightly critical, enough to lower your self esteem and hence make you more vulnerable to buying crap from Amazon.
This seems like it is designed mostly not for the stated purpose, but rather to solve some data collection problem that Amazon has. Note the emphasis on full-length photos and photos from all sides including the back, something they couldn't get simply by scraping, say, http://lookbook.nu, because normally most people in sites like that don't take photos of the back.
Indeed - it would work great for training a neural net for recognizing the human form (or specific humans' forms) from any angle, instead of relying on conventional social media images which seem heavily front-oriented. Now when you happen to be walking away from the camera or trying to hide your face, ML will recognize you anyway!
Crystal ball: this is the Tesla self-driving car strategy. First deploy the hardware, then figure out the software.
In this case the real point of the depth-sensing camera hardware and machine-learning image analysis isn't hands-free selfies and robo-fashion advice. It's building a 3d model of your body to show what you would look like in different clothing, creating a clothes shopping experience that is competitive with shopping in person, and letting Amazon take on a big chunk of retail where it's not currently an appealing option.
100% correct. There is no way they included a depth sensor in this device just to blur the background. They may also be aiming to map any Amazon clothing purchases to 3D body shapes for a purely data driven approach to clothing recommendation to start with (as opposed to having 3D scans of every bit of clothing to do geometric fitting, which is prohibitively labor intensive.) So they'll know your shape, what exact items you've purchased, how often you wear those purchases, and (if I understood correctly) how your friends think you look in them.
Don't forget collecting a huge number of really well-framed, massively-metadata-tagged photos of peoples' otherwise-relatively-private bedrooms and closets.
I look forward to being denied a loan because people in my neighborhood have displayed increasingly-messy closets, indicating growing stress and probable imminent economic trouble among my neighbors, and they couldn't find any spy-photos of my closet so they're very sorry, but they can't make an exception because they can't second-guess the machine, only feed it more data.
Or, you know, all kinds of other horrible dark-future stuff that's increasingly part of our reality. Aside from the usual of making their constant psychological warfare (targeted advertising and marketing) more effective, which is bad enough.
I am not the target customer. However, I really want to show this idea off to a bunch of people and see what they think.
People want to look good and having an automated system that can give you a "once over" seems nifty. Yes, it sounds vapid and for insecure people, but people waste time on being anxious about how they present themselves.
Why is everyone saying this is for insecure people? I'm also not the target audience for this, but I don't think people who are concerned about the way they look are inherently insecure.
I am not insecure (I think !?!?) but it would be nifty to have Alexa tell me that "harshaw, those jeans you are wearing have a giant stain and you should get a clean pair out of the drawer".
Seems pretty clear to me that this is just an intitial step. Echo Look as is may end up being very popular in its own right, but the end goal is probably to move beyond evaluating your existing wardrobe, and eventually making recommendations for clothing sold thru Amazon that you can try on virtually right at home. We've had that for glasses for a while, and believe some startups have already tried this for clothing.
Retail stores are shuttering at an increasing rate as things move online. But buying clothes online requires supporting lots of returns. Imagine you work at Amazon. You have an insane amount of computing power and very capable big data teams. You are tasked with building up clothing sales and minimizing returns. AR is on the horizon but not here yet.
So what could you release today to move towards your goals?
This seems like a solution looking for a problem. I guess an always-on internet microphone wasn't a big enough invasion of privacy, we need a camera too.
>This seems like a solution looking for a problem.
That would characterize most of consumer IoT. Industrial IoT is more interesting, but who needs their bathroom mirror ordering more razorblades for them. All sorts of problems with that. Yet it's nevertheless being pushed by big co's desperate for new revenue streams.
So now Amazon (or whoever is in control of the device) not only has access to voice recordings of everything that goes on in peoples homes, but also video. What could possibly go wrong?
And remember that this is by the company that has a 600 million dollar contract with the CIA[0]. Or in more accurate terms: the known upfront amount is 600 million.
I was initially a little bit confused by Amazon's attempt to disrupt the mirror industry (and it seems like others in this thread are too). After reading through everything, my best guess for the motivation is:
> Echo Look helps you discover new brands and styles inspired by your lookbook
This feels like a marketing vehicle for high-end clothing and accessories. Maybe it's a way to leverage the reach of their Alexa ecosystem to compete with the likes of Polyvore?
Okay, so even privacy concerns aside and assuming this works as well as they say, is there anyone who would actually find this sort of thing useful?
I mean, I'm normally pretty open-minded about products like this, but I'm having trouble seeing the target market here. Are there really a significant number of people out there who want to be able to browse their wardrobe on their phones and get a computer to help them choose their outfit? (I honestly don't know; fashion doesn't really interest me so I'm clearly outside their target market.)
I wonder if Amazon's real motivation for this product may actually be to create a better way to sell clothes online. Maybe once they get enough data points they can start using AR to show you how you'd look wearing a particular outfit or piece of clothing sold on amazon.com? Or maybe it will take your measurements to help ensure the clothes you're ordering will fit you properly. If so that might actually be useful; and would probably significantly increase online sales of clothing.
I would never buy an Echo because of privacy concerns but this does appeal to me.
I don't have fashion sense. And I know something as simple as an algorithm telling me if the colors match would improve how I dress.
That said I'd probably only use something like this if it worked using my existing phone or desktop. Definitely wouldn't want an always on device specifically for it.
Male and female engineers are expected to avoid an appearance of being concerned about looks. Lots of outside hobbies are accepted in the engineering community, but not if your hobby is fashion. It shouldn't be this way, but it is.
Things that I value when purchasing and choosing clothes:
* I want to avoid lots of decision-making and info gathering.
* I want it to be simple to buy.
* I want not super expensive, somewhat stylish, somewhat innocuous, occasionally things that I find unique and cool.
This seems like it would make all that easier. Probably will not get one because I don't care enough about the whole area, but I could see buying it if I wanted to improve my personal style for whatever reason or was more focused on being fashionable/stylish than I am at the moment.
Basically, Spotify Discovery Weekly for clothes. I doubt it will take off as fast as the Echo and it has more flop potential, because "Semi-weird Camera in your Closet" isn't as established a product category as "Radio in your Kitchen" but I think it could take off and also not be some horrible distopian thing.
An algorithm recommends easy-to-buy clothing for you based on a picture and your personal taste... Okay cool!
Remote Worker: Alexa, how does today's t-shirt and pajamas ensemble compare to yesterday's t-shirt and pajamas ensemble?
Remote Worker: Given these two photos of me in different shirts, but the same pajamas, which is more in style, and which is a better fit for me, given my personal style?
What I see here is a bunch of smart tech-savvy people that never took a #lookoftheday pic and posted on Instagram and will never gonna buy the product anyway talking about privacy. How do you know you are completely safe with your phone, with your laptop?
Fact is that there are millions of people posting "look of the day" on Instagram, Snapchat every day. Taking selfies in front of the mirror in their bedroom and they just don't care. I bet they care about a product that would make that easier and afaic Echo Look is promising to deliver that. So I'm curious to see how the actual target will respond to the product.
They seem to imply that humans may see your Style Check photos (ostensibly "our team of experienced fashion specialists" for machine learning purposes). I'm wondering a bit if that team of fashion specialists is Mechanical Turk?
I'm wondering if your picture will appear in the amazon ads if Alexa recognizes that new shirt as a recent order.
Also given that all the pixs in the ad are women, we can assume if Alexa sees a little baby bulge and the other signs of pregnancy that your amazon advertisements are about to take a huge swing to the maternity style.
I definitely searched around to see if this was some late April fools thing, or something like that. I just completely and totally do not understand this product. Plus all the privacy concerns of it...
I'm fascinated by the fact that Amazon clearly did enough market research to think that this will go down well, yet I wouldn't personally consider allowing one in my home.
Feel like the people who would even desire this sort of fashion feedback (young -> college-aged women) would rather use $200 to buy additional items to wear. Much older than that and people have either developed a sense of personal style or have decided they just don't care about clothes. Weird product.
>"Sensors plus machine learning replace keyboard/mouse/touch screen. Camera/microphone as control surfaces."
Feel consistently SV/VC types are overhyping and misrepresenting what is actually going on with products like this. The "Machine Learning" element of this is basically just speech detection (as is the case with almost all these assistant products) used as the equivalent of a wireless button or switch press.
Even in the video we still see users having to swipe away at screens to do anything beyond the button press. Feel SV/VC types are constantly misrepresenting these button presses as the coming of Iron Mans Jarvis when the reality is they're elaborate light clappers
I do think this is an interesting product but to hail it as machine learning replacing UI is disingenuous
It seems so obvious why hasn't Google Now done it already?
Is it because of security concerns that a home device can "spy" on you at all times in your own home?
By the way, there's a new marketing approach that's been trending lately: Instead of saying "this is a Alexa with a camera", they say what it's for: "look your best".
Eventhough we all know adding a camera can do a lot more than just checking if you look good in these clothes or not (ie security monitoring). Apple Watch has been using the same technique lately. It's been focusing the Apple Watch functionality mainly on sports, eventough it can do a lot more than just sports.
Are you guys seeing the trend? Are you going to use it for your own products and services?
Beyond the basic motion detection where you can define areas of interest in the Nest camera's field of view and it will notify you when it detects movement. They've started using machine learning with them so it can distinguish people and notify you when it sees a person. It is also pretty good at not notifying you of false positives. I have one watching my front yard and early on it would notify you of motion from cloud shadows or wind shaking trees but it seems to have learned to ignore them now.
I look forward to them being able to identify UPS, FexEx and USPS delivery vehicles dropping off packages.
From a privacy perspective, previously you had to be worried about Alexa listening to your conversation, now you have to worry about it seeing and listening you.
I am not saying that Amazon is snooping on its users. All i am saying is that such devices are a prime target for hackers and government agencies. Recently wikileaks showed that CIA was using Samsung Smart TVs for snooping on its users[1]. Devices like this can are very attractive targets for government agencies.
I think they are trying to find ways to reduce returns on clothing here. Crazy high percentages of clothing get returned... and why not? To be competitive with retail stores, online stores have to offer free shipping and returns. Try it on, if you don't like it... send it back. Very costly to the seller.
Having a bot tell you, "You are pear-shaped, don't use pants with pockets..." or "This style isn't age appropriate..." -- who knows how complex this can grow. But to start just focus on simple rules every stylist knows; this should help cut back on people buying stuff that looked good on the model, but that they are statistically more likely to have to return.
I had a friend once who walked me through all this while she was shopping. It was interesting to me that there could be, essentially, an algorithm to define what patterns people should use to select clothing. I remember it really wasn't all that complicated... but I'm sure it's a bit more nuanced than the first Google result explains. Anyway I'm pretty confident Amazon will get it mostly right, and be able to grow into the space from there. I think it's pretty cool they are trying this... dressing better is something we can all do. (=
What I see here is a bunch of smart tech-savvy people that never took a #lookoftheday pic and posted on Instagram and will never gonna buy the product anyway talking about privacy. How do you know you are completely safe with your phone, with your laptop?
Fact is that there are millions of people posting "look of the day" on Instagram, Snapchat every day. Taking selfies in front of the mirror in their bedrooms and they just don't care. I bet they care about a product that would make that easier and afaic Echo Look is promising to deliver that. So I'm curious to see how the actual target will respond to the product.
Is Amazon even a serious competitor in fashion in the US? Here in Germany I think they are dominated by other retailers as far as I can tell (it's about the only vertical I can think of where that is the case)
I cringe at the thought of the "modern household" with Dash buttons scattered around the house, Alexas/Echoes in a few rooms, a "Look Assistant" on every dresser, and a slew of other IoT gimmicks fastened on walls and countertops, around every ankle and wrist.
It's straight up tech addiction.
None of it is even impressive. The technology isn't there. And when the technology does get there, I'd rather have it all available from a single device I carry in my pocket, not across a bunch of gadgets that nickel and dime away my attention.
I could see this being a great asset in a department store. Outside of the privacy stalls in the dressing area, that is, and supplementing the part where you walk out in a prospective outfit, get Look'ed, and then use the store iPad to view the photos/video/panorama.
But I don't see the need for home use, at least with the tradeoff in privacy versus what I can already do with my bedroom mirror. I'm not usually making purchasing decisions when getting dressed for the weekday.
While I'm not a fan of Amazon controlling a webcam in my bedroom it would be nice to tell my SO to ask Alexa the next time she asks for wardrobe advice.
I kinda wonder how well this will work. My experience with the echo is that it understands my voice and questions very poorly, far worse than Google's assistant on phones. I've never tried google home.
I guess identifying a person in a photo and checking if colors and patterns are "in" is easier than NLP. Certainly won't have feedback on fit though.
It's a useful source of data, but consumers who invest in high fashion typically post looks on Instagram, often daily, that are littered with phototags identifying brands. The data is out there...
I really like how this product has SO MANY use cases but they have picked just one to go to market with. More will come later. But for now, it just helps you pick you clothes.
This is good because it keeps you expectations low so you will be pleasantly surprised when it does something else.
They did this with the Echo as well. It was just an egg timer at first and played music.
Of course the obvious thing we all did was freak out about privacy, etc. They'll soon be pulling video from the crime scene.
Anyway, can some of the HN crowd take the sunny day case (assume we'll properly address privacy), and discuss interesting uses of this technology? Image recognition and shopping? Show it want you want to buy, for example.
OK, see, I can actually understand the utility of this and it seems like a good idea! As opposed to just the plain Alexa, which seems quite a bit less useful. Very interesting for Amazon to go after the style/fashion set. I mean it still has all the same privacy implications, I'm just speaking of the actual utility of the thing.
"That looks pretty neat. I wonder how well the styling advice part works, but the camera is cool. Fashion bloggers would probably really like it. There are so many people that do a “outfit of the day” type thing"
This device and use-case are fine and fulfill a real need. The hidden cost of using it is moving the overton window on privacy another notch closer to "panopticon".
Too bad it's $200, otherwise it might have had some potential as an inexpensive depth camera to tinker with (although there isn't much detail regarding how "depth aware" it is compared to something like a RealSense camera or a Kinect).
I'm willing to bet this is their first step towards eventually being able to show you how outfits will look on you without needing to try them on. Thus eliminating one of the last remaining reasons people buy clothes at brick and mortar stores.
One possibility is to enforce opinions like yours. In the light of this all of a sudden Echo isn't as strange in comparison. It's a way of normalizing Echo by releasing something more "outrageous".
I would imagine the end goal is to incorporate this kind of tech[0] in the future, thus eliminating the need to in-person shop for clothes. "Depth-sensing" camera? Has to be for that..
With the Alexa they were able to get around the always on monitoring . Not sure about this device but it feels like that Amazon is making things palatable to the consumer like . It's like a consumerist Brave New World.
Will add this video to the list of videos with overly "positive" music that grossly underestimate peoples suspicion. Others on that list: google glasses release advert, apple developers wwdc 2014 intro video
Story:
As an engineer that enjoys fashion but has limited time for the hobby I would like an app that can help me self-evaluate fashion choices and arrangements so that I can have more fun and look better.
Super, we can add "appropriate dress" to our "citizen scores." No more lime green spandex tube tops at the grocery store, or your credit rating will go down.
I've recently built a Slack Doorbell using a Cheap D-Link Webcam and a AWS IoT Button. The one thing that surprised me was the lack of a good internet enabled camera, the D-Link basically has basic AUTH and for capturing photos is the same username and password as for remote. Does anyone know of a good and secure RESTFUL Internet Connected camera?
This feels like a device literature professors would use to teach dystopian fiction writing.
edit: case in point - matthewmcg's comment https://news.ycombinator.com/item?id=14203772