Facebook Delays Home-Speaker Unveil Amid Data Crisis
bloomberg.com
bloomberg.com
However, someone here on HN recently posted the open source voice assistant project mycroft:
Fun to mess around with on Raspi and eventually I may link it up with Home Asssistant.
On first glance, hmmm: why does this open source AI require that I create an account on their webpage before trying it out?
> Mycroft AI, Inc. maintains a device and account management system known as Mycroft Home. Developers may sign up at: https://home.mycroft.ai
By default, mycroft-core is configured to use Home. By saying "Hey Mycroft, pair my device" (or any other request verbal request) you will be informed that your device needs to be paired. Mycroft will speak a 6-digit code which you can entered into the pairing page within the Mycroft Home site.
[...]
> If you do not wish to use the Mycroft Home service, you may insert your own API keys into the configuration files listed below in configuration.
Bingo I'm in! Thanks a lot. This looks awesome.
First, there is no premium version. The whole platform is free to use. Forever. And we are starting to open-source it.
For the trigger words, I think there is a slight confusion about what the are.
There are actually two kinds of trigger words:
- ones which are universal, and are very hard to train (it takes a lot of data and use NN models). Those are paid because we need to get the data (this is the only reason)
- ones which are much simpler and can be trained on your device. This will come in the next release of the platform!
Stay tuned for the next release where the personal hotword is coming!
Which one? And why would you not consider Facebook's speaker but are apparently using somebody else's same product.
https://learn.adafruit.com/using-an-infrared-library/hardwar...
I do something similar, I have an ESP8266 microcontroller hooked up to an IR sender/receiver that I can use to send and receive codes from Home Assistant to control some stuff. The module I use takes care of encoding and decoding the IR and simply dumps it to serial, which the micro reads and forwards to Home Assistant via MQTT (and vice versa). It's this one
https://www.aliexpress.com/item/5V-IR-Infrared-Remote-Decode...
but it's not hard to do it with a bare IR emitter and receiver. Unfortunately I haven't had have much time lately to document many of my projects for other people, but you can learn a lot from Youtube and Google these days.
I use an ultrasonic sensor to detect when the door is open or closed. It's very reliable.
I even wired up Home Assistant to contact me (over sms by sending an email to my number@carrier) if the door has been open for >5 minutes, because I was precisely in your position of forgetting to close it.
With a little z-wave door sensor (just a magnetic reed switch), you could have Home Assistant run an automation that says "If gate is open at 9:30am, close it." so then you don't even have to log in to click it.
[1] https://partofthething.com/thoughts/adding-a-ir-send-and-rec...
Some have a setting to close a few minutes after being opened. Might be simplest.
I've been wanting to add this to my setup. Do you have any more details on what you have done? My main hurdles would be choosing decent cameras that can be used by Home Assistant... Cameras that don't want to phone home all the damn time.
What camera?! Seriously though, I don't necessarily want to write device drivers for a camera, maybe it's time I start looking at usb webcams to use?
There are also a ton of different 3D-printable mounts and jigs for them on sites like thingiverse and instructables.
My favorite[1] comes as a kit with the jig, 3 servo motors, and an add-on board for a raspi. Using this I have a live-feed camera that can rotate up to 180 degress on 3 axis that I can control from anywhere in the world using my phone as long as my internet connection is up at home. All hosted locally, even the live feed frontend. It's not hard to build or set up either, took me a lazy day and I'm far from a good coder.
1: https://shop.pimoroni.com/collections/raspberry-pi/products/...
I remembered incorrectly, that PiMoroni kit does not include the 3 servo motors, you need to buy those separately along with the RPi camera module. Luckily the servos required are super common little blue servos that are all over eBay and such for dirt cheap.
[1] https://partofthething.com/thoughts/raspberry-pi-home-assist...
[2] https://www.home-assistant.io/components/camera.amcrest/
https://medium.com/snips-ai/snips-nlu-is-an-open-source-priv...
The whole platform runs on-device which makes it ideal for privacy, cost, and to allow it to run when there is no network, from hotword to ASR and NLU
We are available in English, French, German, and soon Japanese and Korean, with more European languages coming!
Take a look at what some people have built with it: https://github.com/snipsco/awesome-snips
and a few tutorials to get you started: https://medium.com/snips-ai/building-a-voice-controlled-home...
I like Spotify but I’d also like to be able play my local music with voice commands.
Looks like it can all be done offline + OSS except for Speech-To-Text. But I think the key idea is even if it's using cloud for processing, you can at least trust it isn't listening when it shouldn't be (as long as always-on, wake word detection is offline).
You can then use any language to create skills to make it perform actions.
Take a look at how to build a connected speaker with Snips: https://medium.com/snips-ai/building-a-voice-controlled-home...
All that said, I believe Apple, Amazon, and Google all have worlds more trust in this area than Facebook.
They're also orthogonal to smartphones.
> why do you think they aren't lying about what their mobile OSes or apps are doing?
Why do you think everybody thinks that? And why frame it that way from the get go, to what the companies themselves are doing today, and/or to what the widgets do in isolation? It's another vector of attack for all sorts of actors, present and future, including the companies themselves.
> All that said, I believe Apple, Amazon, and Google all have worlds more trust in this area than Facebook.
They're not the baseline, the baseline is none of these devices. And for me, those all have worlds less trust than entities I trust, and the difference between them and Facebook is pretty much irrelevant in the big picture.
I don’t like the 1 second of buffer the iPhone has for Siri with Hey Siri disabled - but LOL if I’m going to install a single purpose always listening microphone for gimmick reasons.
I disagree.
> they ride off your own network so it is easy to see when they are sending data back home unlike phones which can transfer data over harder to monitor 3rd party networks
…if you use a VPN, then you can check for traffic here as well.
> They are cheaper and lower powered machines so they can't do a lot of processing on device and can't store a large amount of data for later transfer.
HomePod has an Apple A8 processor, and Google Home has some sort of ARM SoC. These aren't low-powered machines; on the contrary, they can run 24/7 since they're always powered.
> They are stationary in your home and therefore likely spend a majority of the time listening to nothing
I think this makes them more likely to be overlooked, more than anything.
> Finally, if you believe that these companies are shady enough that they are lying to their customers about what these smart speakers are doing, why do you think they aren't lying about what their mobile OSes or apps are doing
Some of them, maybe.
This is conditional on the assumption that the device is doing what it tells you it is doing. Like the point in my previous comment, if you believe the device is lying to you about when it is recording, why do you trust that it isn't hiding non-VPNed connections from you? Smart speakers can not physically transfer data themselves and can only forward data over WiFi or Bluetooth.
>HomePod has an Apple A8 processor, and Google Home has some sort of ARM SoC. These aren't low-powered machines; on the contrary, they can run 24/7 since they're always powered.
The A8 is a 4 year old processor and the HomePod is by far the most expensive and powerful of these devices. Maybe I was downplaying their potential processing power too much, but the fact still stands that a modern smartphone is much more powerful than a modern smart speaker.
>I think this makes them more likely to be overlooked, more than anything.
And phones aren't overlooked when they spend almost 24/7 within a few feet of the owner?
I am not saying these devices don't present any potential issues. I am just pointing out that the risk is lower than the one we have already accepted by using smartphones. Stressing over compromised smart speakers is like stressing about whether you locked the door to the third floor balcony while not caring that the front door is unlocked.
Ahh, so you don't trust the operating system itself, rather than the apps running on it. In this case, you're free to physically turn off cellular data on your phone (e.g. by removing the SIM) and connect it to Wi-Fi.
> The A8 is a 4 year old processor and the HomePod is by far the most expensive and powerful of these devices. Maybe I was downplaying their potential processing power too much, but the fact still stands that a modern smartphone is much more powerful than a modern smart speaker.
Sure, but you don't need a whole lot of processor power. What you need is a reliable power source, which is something that a smart speaker has.
> And phones aren't overlooked when they spend almost 24/7 within a few feet of the owner
While smart speakers don't have to even be within a few feet of the owner to work.
The argument against any of these devices is that they are compromised either by a third party or by the actual device maker. Google and Apple make the OSes installed on almost every mobile device. If you don't trust Google's or Apple's smart speaker why do you trust their OS?
>In this case, you're free to physically turn off cellular data on your phone (e.g. by removing the SIM) and connect it to Wi-Fi.
That isn't a valid solution became it is hindering a device in such a way that it can't perform its most basic duty. Real people don't put a SIM into their phone only when they are expecting a call and take it out as soon as the call is over.
>Sure, but you don't need a whole lot of processor power. What you need is a reliable power source, which is something that a smart speaker has.
What do you think the average smartphone uptime is per day? I am willing to bet it is approaching 24/7. A majority of phones probably are hooked up to chargers a couple hours a day and they all have access to the battery any time they aren't/
>While smart speakers don't have to even be within a few feet of the owner to work.
And neither do phones. The range of smart speaker microphones is likely within an order of magnitude of the range of smartphone speakers but the average distance between a smart speaker and its owner is going to be several orders of magnitude higher than the average distance between a smartphone and its owner.
This isn't intended as a permanent thing: it's just a spot check on the traffic.
> A majority of phones probably are hooked up to chargers a couple hours a day and they all have access to the battery any time they aren't
The issue here is that any sort of processing is very noticeable on smartphones, since it will either cause the battery to deplete very quickly or charging to take a long time. It's basically impossible to get away with performing computation without it showing up in battery statistics.
Yes the processor is still strong but it's weaker than smartphones. Yes you can still sneak data past home networks but it's harder than 3rd party networks. Yes they can be more overlooked, but they still have access to less data than smartphones.
None of the points you make prove that home speakers are as bad or worse than smartphones. The original comment wasn't arguing that smart speakers are safe, they were just pointing out the hypocrisy over owning a smartphone and being against smart speakers.
And? Your phone is with you 24/7
> they have access to continuous power
I don't know about you but my phone hasn't run out of power in months.
> this doesn't mean that they aren't less dangerous
No one is saying that they aren't dangerous, you're missing the point. What I'm saying is if you carry a smartphone around, that's just as risky. Therefore, it's hypocritical to warn against smart homes unless you also don't have a smartphone.
> Smart speakers also have better range than smartphones
But smartphones are literally next to you at all time, they don't need better microphones.
I don't understand your argument here. If your smart speaker 100% maxed out its CPU and sensors for a month, would you notice? Probably not. I guarantee you would if your smartphone did, though, unless you keep it constantly plugged in.
> But smartphones are literally next to you at all time, they don't need better microphones.
I'll chalk this one up to individual preference. Personally, my phone is usually not near me at home (e.g. downstairs) since I have access to my computer, which I prefer using instead.
It's not hypocrisy to warn of dangers, regardless of being exposed in other ways. Also, your comment here equates 'smart speakers' with 'smart homes', which are not the same thing (yet).
This conversation can only progress when you consider the actual threat models (like security folks talk about). It seem strange to me to rant about the risks of smartphones vs smart home devices, when the data from both of those is potentially ending up with the same small set of companies.
Smart speakers are sending the audio to the mothership by design, it's the way they work in normal operation.
These things send back full audio to be processed by design. So, it's not a question of possible it's up to the provider what they want to do with the recordings the device is directly sending.
But if you think there will be people willing to self host and maintain a whole voice assistant service, you're off your rocker.
It's just like saying "Why doesn't everyone just use Tor and PGP?" Because it's an enormous pain in the ass and really only accessible to technical users.
There were probably people who said that about open source operating systems, once upon a time.
Extremely complicated software is packaged down to an installer and installed by non-expert users all the time.
But moreover, if you need to call me “off my rocker,” you’re not engaging in good faith.
Nobody reads TOS.
Car analogy: leaving you car in a bad neighborhood, unlocked, key in the ignition, door open, with a "do not steal or we will send you a threatening letter demanding our car back" note.
I'd suggest you turn of all the personalization ads option they provided so that you can avoid to be precisely targeted
> When we do send information to a server, we protect your privacy by using anonymized rotating identifiers so that searches and locations can’t be traced to you personally.
Additionally, I can disable the microphone on my home device if I wish with the push of a button. If I'm in public I can't walk around asking every stranger within earshot to "disable their phone's microphone".
Now, explain to me, in the case where "Ok Google" is enabled; how isn't the phone always recording if "Ok Google" works (or "Hey Siri" for that matter)?
Given everything covered, is it still "disingenuous" to say that a cell phone is minimally as risky as putting any of the devices mentioned in a home if not more due to the inability to control the devices of others?
Facebook was called Home. And it had some nonsense gimmick with a Facebook-like Android skin for the OS.
But, yes it failed on its own merits when the consumer market took one look and laughed it off.
edit: I remembered wrong. Home was the name of their Android skin. HTC and Samsung launched a few phones running it by default, which all flopped.
There's not much difference between having the Facebook app installed on your phone and this device, in terms of creepiness and data collection capability. Probably their smartphone app creeps me out more, actually.
On my install Facebook has zero permissions and remains useable. It is by design though.. they have already way too much data.
IIRC in P only a foreground app can access sensors, still, don't give this permission to facebook.
Sure, every regulator would bark, but that's kind of exactly how Uber did it for ridesharing!
People flying around with probably alcoholic tendencies, overly adventurous ideas, spraying toxic insectesides, or otherwise borderline pilots because "gig economy". You're probably right.
Anyway, not much blood to squeeze out of that stone. A 747 fits about 400 passengers and only needs 2 pilots. Doubling or halving the pilot salary will not make a noticable difference in ticket prices.
I'm not sure what the main costs of air travel are, but unlike taxis, it ain't the drivers. I would guess fuel?
It needs many more than that. Depending on the workload it could be 10 or more per plane.
Consider a daily shuttle between LAX-NRT, a 11h40m flight. This will usually require 3 pilots, as pilots cannot legally fly more than 8-9 hours per day. When they arrive, the plane is turned around in a few hours and comes back - but the original crew needs rest, so at least 3 pilots have to be waiting at the other end, ready to go. And when the plane arrives back in LAX, who's going to take it back to tokyo? Those 3 are resting! So straight up that's 9 required, at an absolute minimum, with no margin for sickness or holiday.
Longer segments can require 4 pilots, so again that's 12 at an absolute minimum. Realistically speaking, depending on a whole bunch of factors, airlines would aim to actually have about 15-20 pilots per plane in a long haul heavy scenario.
And we haven't even mentioned cabin crew, which would be 12 or so on our hypothetical 747. So again, multiply that by 3 at a minimum and you've got something like 40 cabin crew per aircraft.
So 15 highly paid pilots (say 150k/yr) and 45 slightly less paid cabin crew (50k/yr) - that's close to $4.5m/yr just on crew. For ONE airplane. You've got to be making $12k/day on that plane just to pay flight staff, let alone ground staff and everything else.
In the world of narrowbodies the economics are even more brutal. If a plane's on the ground it's not making money, so say you try to have the plane in the air 10 hours a day. Due to flying limits over extended periods you would need 8 pilots to fill that with no margin, and probably 8 or so crew for a 737 with 4 active. Realistically you're looking at a crew bill of $2m/yr, or ~$35 per seat per day. Even if you manage to squeeze 7 segments out of the plane per day with 100% load factor that's $5 per ticket just on crew in a market where every single dollar counts.
Now obviously I'm grossly oversimplifying and pretty much every number I mentioned can vary - but not by all that much. I'd say those numbers are at least in the ballpark for a full service carrier and if anything the long haul estimate is probably too low - I've heard up to 24 pilots per plane for long haul carriers like the ME3. So finally the point is, crew wages absolutely make a difference!
It's a pretty interesting industry.
It's just that "minimally qualified pilots" are still quite highly trained. But the pay for new entrants is quite low relative to the responsibility.
Not a joke...
Public: "Uh we're having some unexpected delays"
The values of Facebook, of which there are none.
Then again why does any one trust the companies who provide this service at all?
I don't think they do. I think either they provide a service that people want, or that they truly have no idea they're the product.
Apple for example takes privacy pretty seriously so I am more than happy to allow their products inside my house.
I don't trust most companies, but that doesn't stop me from buying their goods or using their services.
That said, I think as popular as Echo devices are, people already have a bit of paranoia about them, and that's before considering the controversy FB is now undergoing. Unless the hardware is best-in-class, hard to see how this will fare any better than Facebook's failed attempt at its own phone with HTC: https://en.wikipedia.org/wiki/HTC_First
I can imagine extremely well the field day fb product managers must have had with the notion of a 24/7 real time data feed of people’s homes.
I was given one as a gift, and found it useful enough after a few days that I was willing to trust that it wasn't recording unless it detected the activation word locally.
And then I walked into the room while it was carrying on a conversation with my 2 year old, asking his name, etc. Unplugged it and plan to destroy it.
https://www.amazon.com/gp/product/B073X5FYVF
It says the app "contains curated stories from the Amazon Rapids app library as well as a selection of Audible short stories". It's pretty good, as it features actors and sound effects. Can't say any of the stories were interesting to me (was hoping for Aesop Fables type stories, which is about all I remember growing up myself).
Funny to see the alleged data abuse giant unravelling with all the hustle and bustle of upper management panic.
Here's the information I gathered from 20 minutes on Facebook - a powerful platform indeed.
The author (Sarah Frier) seems to be on good terms with the CXOs of Facebook:
CPO Chris Cox https://www.bloomberg.com/features/2016-facebook-reactions-c...
COO Sheryl Sandberg https://www.bloomberg.com/news/features/2017-04-27/how-shery...
CEO Mark Zuckerberg https://www.bloomberg.com/features/2016-facebook-virtual-rea...
To drive it all home: https://www.bloomberg.com/news/articles/2016-11-22/how-fake-... https://www.bloomberg.com/news/features/2017-09-21/mark-zuck... http://www.bloomberg.com/news/articles/2016-11-11/facebook-s...
For the most part, agencies are skeptical many people are going to leave Facebook - or at least enough to make a difference for advertisers. The public's attention span is fleeting, noted one executive, and even massive data breaches that affected Yahoo, which included 3 billion accounts, and Equifax - 147.9 million accounts - haven't turned people off those services.
...
In addition, the public already held Facebook in low regard leading up to the scandal. According to app analytics firm Sensor Tower, sentiment about the company during the the two weeks leading up to the Cambridge Analytica stories was already 83 percent negative - and people were still on the platform."
Source: https://www.cnbc.com/2018/03/22/facebook-cambridge-analytica...
"3. The idea that users will abandon Facebook is unlikely. There's really no alternatives. An abandonment of Facebook and related apps would basically be an abandonment of social media altogether, which is an unrealistic expectation.
...
4. While Facebook currently produces the majority of its revenue from advertising, that's only the tip of the iceberg. Facebook can do many things to begin diversifying its revenue stream including entertainment or e-commerce.
...
So far Wall Street's opinion on Facebook has not been altered by the data scandal. According to MarketWatch, 37 out 44 analysts rank the stock as a "buy". The average price target is $221.56, which represents 39% upside based on the current stock price of $159.09."
Source: https://seekingalpha.com/article/4158807-facebook-let-others...
There's also bubbly-market effect, and these might result in deflation a bit.
The public wasn't turned off of Equifax because they weren't Equifax's customers in the first place.
The public aren't even Equifax's users.
American equity "analysts’ recommendations...provide no (or biased) information related to future stock performance (e.g., Jegadeesh, Kim, Krische, and Lee 2006; Drake, Rees, and Swanson 2009...Bradshaw 2009)" [1]. Some research even concludes "that stock recommendations in the U.S. are either insignificantly (Bradshaw 2004) or negatively (Barniv et al. 2009; Drake et al. 2009) related to future returns" (page 4).
Meanwhile, from the 16 March to 27 March (close to close), Facebook dropped 18% [2]. Over the same interval, the S&P 500 lost 5.1% [3]. (Facebook is in the S&P 500 [4]; the rest of the S&P 500 did better than -5.1%. I chose 16 March since that's the day before the Cambridge Analytica breach broke [5].)
[1] https://pdfs.semanticscholar.org/b4e8/ffa418d387fb57a7cabe89... page 2
[2] https://finance.yahoo.com/quote/FB/history?p=FB
[3] https://finance.yahoo.com/quote/%5EGSPC/history?p=%5EGSPC
[4] http://money.cnn.com/2013/12/11/investing/facebook-sp-500/in...
[5] https://www.nytimes.com/2018/03/17/us/politics/cambridge-ana...
Really? My parents and siblings, and many friends, tried it, decided it wasn't all that interesting, and left. That's anecdata, but it spans several demographics. People were fine before "social media," so they can probably muddle through without it.
> Facebook can do many things to begin diversifying its revenue stream...
This is where things get scary. Facebook is a surveillance company, and when they can't make money with ads, they will make money in more nefarious ways.
> current stock price of $159.09
Make that $152.22. They're still searching for the bottom.
This assumes that those persons are 1. using browsers that automatically load images, even ones that serve them no purpose and exist only to feed data back to FB and other third parties and 2. taking no steps to prevent their computers from accessing these Facebook domains.