Amazon is Developing an In-House AI Chip to make Amazon Alexa Faster

Amazon is Developing an In-House AI Chip to make Amazon Alexa Faster

Amazon is Developing an In-House AI Chip to make Amazon Alexa Faster

For years, Google has called themselves a mobile-first company as they were transitioning their systems to meet the growing trend of smartphones. These mobile devices are still incredibly important to the growth of Google (as well as other websites), but then the company shifted focus towards AI and now call themselves an AI-first company. This shows you how important artificial intelligence will be in our electronics over the next 5+ years, so it’s no surprise that a new report says that Amazon, one of Google’s fiercest competitors, is currently designing an AI chip to improve Alexa.

Amazon continues to dominate the smart speaker market, but Google’s market share is growing fast. We have yet to see how Samsung’s Bixby smart speaker will perform and whether or not Microsoft will get serious about smart speakers powered by Cortana. Apple recently released their product, the HomePod, but it’s so expensive that very few people will be able to outfit every room in their house with one. So while all of these products do similar tasks, the winner will be the one who performs those tasks the best and has extensive third-party support.

Unique features is also something that could give one company the edge in the newly competitive market. According to a report from The Information, that new feature will be the integration of artificial intelligence. Amazon acquired chip designer Annapurna Labs back in 2015, so that could be the team working on this new chip. The report says the goal of having this A.I. chip in their plethora of Echo devices, and other hardware powered by Amazon’s Alexa virtual assistant, is to make Alexa respond more quickly.

With these Amazon Echo products powered by Amazon’s in-house AI chip, Alexa will be able to handle certain tasks without ever sending a request to the cloud for processing. This will result in faster response times for some of the most basic and common voice commands.

Source: The Information

Want more posts like this delivered to your inbox? Enter your email to be subscribed to our newsletter.

via xda-developers

February 12, 2018 at 10:30PM

Apples Homepod ist der dümmste aller smarten Lautsprecher

Apples Homepod ist der dümmste aller smarten Lautsprecher

Homepod. (Bild: Apple)

| Hinweis: Artikel enthält Provisions-Links – Was ist das?

Der Homepod, Apples erster smarter Lautsprecher, besitzt lediglich reduzierte Siri-Funktionen. In einem umfangreichen Test schneidet der Speaker merklich schlechter ab als die Konkurrenz und ist damit der „dümmste“ smarte Speaker auf dem Markt. 

In einer Untersuchung hat Loup Ventures, eine Venture-Capital-Firma mit Fokus auf Forschung, die smarten Lautsprecher der großen Hersteller verglichen. Dabei kommen sie zu dem recht eindeutigen Urteil, dass der Homepod deutlich „dümmer“ als die Konkurrenz ist.

Homepod: Letzter Platz bei smarten Funktionen

Loup Ventures zufolge konnte Siri auf dem Homepod-Lautsprecher gerade einmal 52,3 Prozent von 782 unterschiedlichen Fragen korrekt beantworten. Die Konkurrenz machte eine signifikant bessere Figur: Amazons Alexa erreichte 64 Prozent und auch Microsofts Cortana erzielte mit 57 Prozent ein besseres Ergebnis. Als eindeutiger Sieger stellte sich allerdings Google Home respektive der Google Assistant heraus: Insgesamt erreichte der Assistent den beachtlichen Wert von 81 Prozent.

Passend zum Thema: Homepod: Was Apples Siri-Lautsprecher kann – und was nicht

Die Tester äußerten abgesehen von der Kritik an den spärlichen Siri-Funktionen auch positive Aspekte: So könne der Klang des Lautsprechers überzeugen. Er wurde gar als der am besten klingende Smart-Speaker bezeichnet. Ebenso gefiel ihnen die User-Experience besser als die der Konkurrenz. Darüber hinaus sei die Spracherkennung der der Konkurrenz überlegen: Im Testzeitraum verstand der Speaker dank der integrierten Geräuschunterdrückung beeindruckende 99,4 Prozent aller Anfragen korrekt – lediglich an den richtigen Antworten müsse Apple noch feilen.

Der Google Assistant dominierte die Untersuchung – der Homepod belegte den letzten Platz. (Grafik: Loup Ventures)

Homepod: Tester sehen Hoffnung in Sachen Siri

Loup Ventures sehe trotz der mauen Siri-Resultate noch Hoffnung für Apples Siri-Lautsprecher. Denn wie bereits erwähnt, ist der digitale Assistent auf dem neuen Produkt um zahlreiche Funktionen beschnitten. Sobald Apple Features wie das Anlegen von Kalendereinträgen und das Verschicken von Mails ermöglicht, könnte der Speaker stattliche 67 Prozent erreichen und sich sogar vor Alexa positionieren.

Apples Homepod kommt in weiß und schwarz. (Foto: Apple)

1 von 11

Laut den Prognosen von Loup Ventures könnte Apples erster smarter Speaker einen ähnlichen Erfolg erzielen wie die Apple Watch. Das Unternehmen geht davon aus, dass Apple in Kalenderjahr 2018 sieben Millionen Homepods – etwa 12 Prozent des globalen Smart-Speaker Marktanteils – verkaufen und damit ein Prozent am Umsatz des iPhone-Konzerns beisteuern könnte.

Auch wenn die erste Genration des Homepods primär darauf abziele, ein toller Lautsprecher mit Siri an Bord zu sein, hege Apple weit größere Absichten mit der neuen Produktsparte, spekuliert Loup Ventures. Langfristiges Ziel sei es, Siri zu einem omnipräsenten, im Hintergrund stets agierenden Dienst zu machen, der alle verbundenen Geräte und Dienste miteinander verbinde und steuere.



Hinweis: Wir haben in diesem Artikel Provisions-Links – auch Affiliate-Links genannt – verwendet
und sie durch „(Provisions-Link)” gekennzeichnet. Erfolgt über diese Links eine Bestellung, erhält eine
Provision vom Verkäufer, für den Besteller entstehen keine Mehrkosten.

via ❤ t3n

February 12, 2018 at 03:04PM

Hygge Home – Alexa Smart Bath

Hygge Home – Alexa Smart Bath


‘It is late December, the days are short, the nights long. Your journey home from work is a bitter cold and dark one, you long for nothing more than to arrive at a home alive with the flicker of candle light, where a hot drink and warm bath await you. Instead, when you open the door, the house is eerily quiet, rooms filled with the winter darkness.’


Hygge (pronounced hue-guh) is a Danish concept used to describe a distinct feeling of cosiness and warmth, it’s extremely prevalent in Danish culture and has recently spread across many homes around the globe.

My project utilises an Arduino controlled Smart Bath to instil a sense of Hygge, creating a homely and warming atmosphere upon your arrival home via Alexa voice integration.

The main control unit for my system is the Arduino Yun Mini, this is used to control the Solenoid valves which control the flow of water.

As you can see in the video above I created a fully functioning prototype of the bath, using a hose to supply pressurised water to the cold side. Not only did this enable me to test the design but I hope it will clearly illustrate how everything operates in terms of the plumbing involved. The bath can be operated both manually or via Alexa just in case the internet ever cuts out and you wanted a bath. The bath fills up on a timer programmed within the Arduino Code, there’s also a reset switch which makes sure you can’t accidentally turn it on twice and overflow the bath.

I created and used an Alexa routine for the video which meant when I say "Alexa, I’m home" the bath turns on and I receive a welcome home greeting. Below is a video to setup your own:

Disclaimer: If you choose to follow this tutorial to create your own smart device you do so at your own risk. I’m neither a trained electrician nor plumber and therefore would not recommend installing these devices in your home without professional assistance. Dealing with mains power is extremely dangerous and can result in major injuries or even death. There are also regulations around the use of powered devices in a bathroom, again I would not recommend installing the bath control without professional assistance as a single code error could lead to a flooded bathroom.

The Skill

You can find the published Alexa skill here or search for "Hygge Home": this gives you an example of what the skill will look like however you won’t be able to operate the device as I can’t have strangers controlling my taps…

How It Works

Below is an interaction diagram and explanation which shows the sequence of events that occur when I say a command such as “Alexa, Turn on the bath”. I’ll be going through each of these steps in this tutorial so you can create your very own Alexa smart device!

Voice Interaction Diagram

1. The Alexa home skill allows me to control the Arduino based device with On/Off functionality. This is what you install on the Alexa app to control your device.

2. AWS Lambda receives the commands from the Alexa skill and sends the relevant data to update the AWS IoT Thing Shadow.

3. AWS IoT hosts a virtual version of the device known as the “Thing Shadow” this updates with the desired state of the device such as “Turn On” or “Turn Off”.

4. The Arduino Yun Mini monitors the MQTT feed of the necessary thing shadow and reacts to the required command by controlling a MOSFET.

Build Tutorial

I will be going through each stage of the process step by step with “checkpoints” along the way so you can check your progress and make sure it’s working as expected before moving on. This should help to iron out any issues early on and make things run as smoothly as possible. The tutorial is focused on the smart bath control but could be adapted to many other devices. If you have any questions please don’t hesitate to ask and I will try my best to help.

The Alexa Skill

We will begin by creating the Alexa skill, the Alexa documentation proved to be an invaluable source of information with regards to this and I strongly suggest reading through the getting started pages prior to creating your own.

1. Developer Account

In order to access the Alexa console, you will first need to create a developer account via this link

2. Alexa Skill Dashboard

Next navigate to the Alexa dashboard and click getting started on the left-hand side under “Alexa Skills Kit”

3. Setup – Skill Information

On the next page, you want to click the “Add New Skill” button which will bring you to the following page.

I’ve highlighted in red the key points which are as follows:

  • Skills Type: Smart Home Skill API – This is used to provide basic On/Off functionality and is the easiest type to get started with.
  • Name: Hygge For this tutorial I suggest using the same name first time round then changing the name in the future once things are operational and you understand how each section links together.
  • Payload Version: v3 (Preferred) At the time of writing this is the latest version but is likely to change in the future, the changes will need to be made to the Lambda function accordingly.


3a. Setup –Interaction Model

Not needed as we are using the Smart Home Skill API which takes care of this for us. However, you need to note down your Application ID. Found below the name of your skill in the top left and looks something like: amzn1.ask.skill.xxxxxxxxx-xxxxxxxx-xxxxxxxxxx

4. Setup –Configuration

It’s in this section where you provide the bulk of the back end information the skill requires, we can fill in most of this however the AWS Lambda Endpoint will have to remain blank for now. This section also requires us to setup some authorisation details which will generate the required security information for us to use.

4a – Security Profile

We need some extra details to be able to authorise our skill, in this section we will setup the appropriate security profile and note the information we need.

First navigate to select “Create a New Security Profile

Then fill in the required information and hit “Next

Now navigate to the security profile page and select the profile you just created. From here make note of your Client ID and Client Secret

4b – Web Settings

Click the “Web Settings” tab then “Edit” and then copy in your Return URL.This can be found on your Alexa Skill configuration page about half way down.It will look something like this:

4c – Configuration

  • Scope: For this we just need to add the following: “profile:_user_id
  • Authorization Grant Type: “Auth Code Grant
  • Client Secret: Paste your client secret from step 4a here.

5. Test

Once that is saved you can now open the Alexa app on your preferred smart phone of choice and download and install your skill! It won’t have any functionality but it’s there and if you can successfully authenticate then it’s working.

AWS Lambda

That’s the skill done for now, hopefully you have the skill installed on your Alexa app. It’s now time to use AWS Lambda to add some device functionality.

1. Create Function

Head to and click the “Create Function” button in Orange. You will notice I’m using the “EU-West”region, I suggest keeping to this for the purpose of this tutorial and changing to suit your own region afterwards.

2. Setup Options

  • Role: Create Custom Role -This will open a new window where you should select the following values:
  • IAM Role: Lambda_Basic_Execution
  • PolicyName: “Create New Policy” then edit the policy and paste in the following:

{ "Version": "2018-12-01", "Statement": [ {"Effect": "Allow", "Action": ["logs:CreateLogGroup", "logs:CreateLogStream","logs:PutLogEvents" ], "Resource":"arn:aws:logs:*:*:*" }, { "Effect": "Allow","Action": "iot:UpdateThingShadow", "Resource": "*"} ] }

Click “Create Function” Bottom Right

AWS Lambda – Create Function

3. Link to Skill

On the left hand side, we need to select the trigger for the function. Select “Alexa Smart Home”

If you scroll down it will ask for Application Id, this can be found on your Alexa Skill developer console. Refer to step 3a.

You can also now copy your Lambda function endpoint into the Alexa skill configuration page.

AWS Lambda – Application ID

4. Code

Once that is added if you scroll down you will get to the code editor. Simply remove any existing code and paste in the Lambda code found below and click save! We will replace the dummy information later to match your own devices, however for now make no changes to the code.

5. Test

The code you have copied in contains dummy data which will need to be replaced by your AWS IoT Thing data once setup. For now, If you search for devices on the Alexa App you should see the device. If this is correct then so far so good.


You should now have a functioning skill and lambda function. Now the next step is to create your virtual “Thing” via AWS IoT. The lambda function updates a virtual device which can then be mimicked by the Arduino via its mqtt feed.

First, we will setup the basic “Thing” then generate the required certificates for the Arduino. It’s important to keep the certificates saved and backed up as you only get one chance to do so.

1. Setup

Head to “”which should bring you to the following page, click “Create a Single Thing”.

1a. Thing Information

On the next page you just need to provide a name for your“Thing”, I’ll be using smart-bath. This is the name you’ll refer to when controlling your device, it’s a good idea to keep this consistent across the different stages for ease of use. Scroll down and click Next.

AWS IOT – Thing Information

1b. Certificates

On this section we need to click “Create Certificate”

AWS IOT – Create Certificate

This will bring you to a page where you need to download and keep note of the following:

  • “A certificate for this thing”

These keys are crucial to the Arduino communication with the Thing Shadow and this is the only chance you will have to note them down

Click “Activate

Then “Done

2. Policy

You should then be redirected to the Thing Dashboard and see the device you just registered. Next we need to create a permission policy for the monitoring. You can either head to Secure> Policy > Create on the left side or visit

Fill in the information as per the example below and click “Create

AWS IOT – Policy Information

3. Attach Policy

The last step is to head to the certificates section on the left hand side. Click the three small dots above your certificate and click “Attach Policy”.

If you click on the certificate you should now see both the Policy and Thing attached to it. The next step is to link this to the Lambda Function so we can update the Thing Shadow via Alexa.

4. Note Thing Information

Before we link the lambda function we need to make a note of a few things. Head to your Thing Dashboard and select the Thing you just created. On the left hand side select “Interact”. From this section note the following:

  • Rest API Endpoint:

5. Linking AWS Lambda and AWS IoT

Simply head back to your Lambda Function and replace the end point with the one you just noted down. Your thing name should still be "smart-bath" so there’s no need to make changes for that.

First Test

You should now have your skill linked to the Alexa app on your smartphone, the skill is linked to the Lambda function and that function can update your AWS IoT Thing Shadow. If you open your thing and select the “Shadow” section you can then turn the device on and off via the app and see this reflected in the shadow document.

If everything is working okay and your Thing is updating accordingly then the main setup of Alexa is finished and it’s on to the Arduino section.

Arduino Setup

The last section is to setup the Arduino to monitor our Thing Shadow and respond accordingly. By stripping out the timing function and reset button this code could be adjusted to suit other devices such as plugs or lamps.

I’m using an Arduino Yun Mini, however the same process applies for the Arduino Yun. You just to need to make sure you have the correct board selected when it comes to uploading the sketch.

1. Arduino Yun Mini –Setup

You may already have done this step but for anyone working with a new board the process is very well outlined here : thanks to Hackster user @BenEagen for finding this guide.


The next step is to set up the SDK for the Arduino board. This is very well documented by Amazon here:

Make sure you use the certificates downloaded in AWS IoT – Step 1b. This will enable secure communication between your Arduino and Amazon services. Once you’ve gone through their guide you will end up at the shadow echo example. Once you’ve filled in your details on the “aws_iot_config.h” you are almost ready to go. It’s worth running the example sketch and opening the serial monitor to make sure you can connect successfully.

3. Smart Bath Sketch

The final step is to simply open the sketch provided below and upload!

The sketch is based on the shadow example given by Amazon. When there is a request to turn on the bath the hot water tap will turn on for the given time, followed by the cold water tap. Once the bath is full the sketch recognises this and will not turn on again until the reset button has been pressed. This is to stop any accidental bath overflows. You can also turn off the bath before its finished, again this will require a press of the reset button afterwards.


That’s all the software completed. If you attach an LED to pin 3 and pin 6 then a button to pin 8 you can test the whole setup. You’ll notice the LED on pin 3 will stay on longer than pin 6. You will need to run a bath manually and time how long each tap is on for to achieve a bath at the desired level and temperature.

Building The Smart Bath

As you can see from the photo below the majority of this is your standard bath tap plumbing and will vary depending on your taps and bathroom setup. It’s key to have the spout and handles separate, as this allows you to add in the Solenoid Tee junction.

As every installation will be different (and I’m no plumber) the dimensions for each section of pipe will vary massively. I used compression pipe fittings for ease of use but soldered joints would work just as well. You will find a list of parts I used above, including the thread sizes needed for each solenoid.

It’s important to use jointing compound on all pipe fittings and some PTFE tape on all threaded connections. I used a hose adapter to test the design under nominal house water pressure.

As i’m no expert on plumbing i recommend the following guide on how to correctly use compression fittings:

Below is also a circuit diagram for the Arduino Yun Mini.

Thank you for taking the time to read this guide! If you have any suggestions on how it can be improved or have any questions then I’ll be happy to hear it and try my best to answer them!

I detailed some of the development on my site :


February 9, 2018 at 07:10AM

Smart Home: Warum uns Alexa nicht abhören kann

Smart Home: Warum uns Alexa nicht abhören kann

An alle Aluhutträger: Die Sprachassistentin Alexa kann euch nicht ausspähen. Das meint ein Nutzer auf reddit und nennt folgende Gründe.


Themen: Amazon, Amazon Alexa, Amazon Echo Show, Echo Spot, Amazon Echo Plus, Amazon Echo, Amazon Fire TV Stick, Amazon Fire HD 8, Amazon Fire HD 10, Amazon Fire Kids Edition, Amazon


January 27, 2018 at 01:42PM

Alexa will one day be able to tell you what it thinks about movies and beer

Alexa will one day be able to tell you what it thinks about movies and beer

Alexa’s favorite beer is Budweiser, and whether or not you agree with it, it’ll one day also have thoughts on TV shows, films, and other subjects. As reported by TechCrunch, Amazon’s voice assistant, which already offers information about what’s showing on TV, will also be able to personally recommend its favorite films, for example. “Having an opinion makes you more interesting, even as an assistant,” Amazon Fire TV VP Marc Whitten told TechCrunch. Whitten states that speaking to an assistant should be like having a conversation rather than just asking fact-based questions. “This is the 2018 version of the video buff at the video rental store,” he said.

While Alexa can respond to questions like “What should I watch?” based on your viewing habits and preferences, these new recommendations instead would be entirely its own opinion. “This is the power of machine learning. One of the most interesting things we’re going at is how do you design an assistant that feels like you’re having a conversation with someone,” Whitten said. TechCrunch reports that Amazon has a separate team working on improving Alexa’s knowledge base by adding more fact-based information and noting the questions that Alexa isn’t able to answer.

Amazon’s long-term goal is for Alexa to be able to have a conversation about her own opinions and preferences without curation from the company. “At the scale we’re talking – Alexa is now launched in seven countries – you can’t editorialize opinion on everything. That doesn’t work,” said Whitten. “The ambitious goal is that you don’t have to do [human curation].”While Alexa might not have decided what its favorite movie is, it can already do some neat things like give you dating advice or let you speak to Destiny 2’s AI companion Ghost.

via The Verge

January 15, 2018 at 12:38PM

Kodi bekommt eigenes MediaThekView-Addon

Kodi bekommt eigenes MediaThekView-Addon

MediathekView war zuletzt Silvester bei uns ein Thema. Die Software erlaubt euch das Durchsuchen der öffentlich-rechtlichen Sendeanstalten, sodass ihr Inhalte später anschauen oder herunterladen könnt. Noch einfacher macht es die Web-Variante, da hier direkt Downloads möglich sind, ohne dass man eine Software installiert. Und obwohl Nutzer von Kodi schon Zugriff auf viele Mediatheken (unter anderem mit dem Mediatheken-Plugin für Kodi) haben, soll es nun noch eine Ecke besser werden.

Die Macher von MediathekView haben die Arbeit eines Entwicklers anerkannt und dieses als „offiziellen Client“ gelistet. Das Open-Source-Addon für Kodi ist derzeit über das private Repository zu beziehen. In Kürze ist davon auszugehen, dass das Addon auch über das offizielle Kodi-Addon Repository erhältlich sein wird.

-> Zum Beitrag Kodi bekommt eigenes MediaThekView-Addon

-> Zum Blog Caschys Blog

via Caschys Blog

January 11, 2018 at 04:17PM

First Alert’s Safe & Sound HomeKit smoke alarm doubles as an AirPlay 2 music speaker

First Alert’s Safe & Sound HomeKit smoke alarm doubles as an AirPlay 2 music speaker

First Alert’s Onelink smart and carbon monoxide alarm (review) was the first to hit the market with support for Apple’s HomeKit, and now First Alert is out with an interesting new version.

Unveiled at CES 2018, Onelink Safe & Sound is a new version that includes a built-in music speaker. The current version includes voice feedback when alarms are triggered, but First Alert says Safe & Sound will actually support AirPlay 2 for real music playback when the feature launches this year.

9to5Mac Happy Hour

Apple’s new AirPlay 2 feature is set to debut with the HomePod sometime in early 2018 then expand to other speakers upon availability. When it launches, AirPlay 2 will enable multi-room audio playback from iOS with improved playback performance compared to the current version of AirPlay, and AirPlay 2 speakers will appear in Apple’s Home app as demoed at WWDC 2017.

Onelink Safe & Sound with Apple HomeKit – Using the Apple Home app or Siri, simply and securely control Onelink Safe & Sound as well as other HomeKit accessories right from an iPhone, iPad or Apple Watch. Set up scenes, access the Onelink Safe & Sound remotely, or have things happen automatically. The Onelink Safe & Sound will also support Apple AirPlay 2 upon availability. AirPlay 2 enables multi-room audio and Siri control for iPhone, iPad and Apple Watch users, all through the Apple Home app.

Using your smoke detector as a music speaker is certainly an interesting concept. My guess is it will come down to speaker quality as to whether or not this feature is actually useful or just an add-on.

Siri, play “The Roof Is on Fire” on the hallway smoke detector.

Safe & Sound doesn’t stop at AirPlay 2 support however. Like other HomeKit accessories, the new alarm also ships with Amazon Alexa built-in for voice control, voice assistant needs, and music playback from sources like Spotify and Pandora. The unit also works with Google Assistant as 9to5Google reports.

First Alert is also entering the networking market with an eero/Netgear Orbi competitor.

In keeping with its reputation for safety and reliability, First Alert enters a new category with its Wi-Fi mesh tri-band solution, which delivers faster, stronger, more reliable internet throughout the home. The Onelink Connect system includes three nodes which can be placed throughout the home. The nodes are in constant communication with one another via a dedicated backhaul, preventing congestion and directing maximum internet speeds to the locations where it’s needed most.

This is CES, however, so pricing and availability are to be determined. You can keep up with all of the news out of Las Vegas with our CES 2018 Guide.

Subscribe to 9to5Mac on YouTube for more Apple news:

via 9to5Mac

January 8, 2018 at 05:02PM

Spectre and Meltdown make anything with chip in it vulnerable, but Raspberry Pi is safe

Spectre and Meltdown make anything with chip in it vulnerable, but Raspberry Pi is safe

Nothing like starting the new year with a security flaw that affects more or less every computer system in the world. Last week we all discovered that pretty much anything with a chip is affected by Spectre and Meltdown, a pair of silicon security flaws that are likely to haunt our steps for a good long time. Basically, anyone with an Intel processor is vulnerable. If you’re running an AMD or ARM processor, you’re likely vulnerable as well.

The only people looking to be safe from this security flaw? Raspberry Pi users, of all things. Those lucky ducks.

What is Spectre? And Meltdown? And why the heck am I vulnerable?

The Register broke the news last week with some very smart reporting about the Intel vulnerability. TechCrunch also has a nice explainer, which you can read here. In essence, both Spectre and Meltdown are security flaws on the chip-level that allow sensitive inside computer systems to be exposed.

Basically, it’s a problem with the chip’s architecture. These bugs are present in all modern Intel processors produced in the past decade.

Meltdown is specific to Intel processors. “It works by breaking through the barrier that prevents applications from accessing arbitrary locations in kernel memory. Segregating and protecting memory spaces prevents applications from accidentally interfering with one another’s data, or malicious software from being able to see and modify it at will. Meltdown makes this fundamental process fundamentally unreliable.”

Spectre affects Intel, AMD, and ARM processors. This means that basically anything with a chip in it is vulnerable, including mobile phones, embedded devices, and more. Spectre essentially tricks applications into accidentally disclosing information that would normally be inaccessible, safe inside their protected memory area. (If you want to read a really excellent layman’s explanation of Spectre, security expert Joe Fitz has a great twitter thread on how it all works.)

So why isn’t Raspberry Pi vulnerable?

Well, as a matter of course, Raspberry Pi runs on an ARM Cortex-A53, making it safe from Meltdown. However, most devices with ARM and AMD cores also use caching and speculative execution, which means they’re vulnerable to Spectre. Raspberry Pi doesn’t, making it one of the few devices on the market that’s free and clear.

The Raspberry Pi blog has an excellent post up explaining exactly how their devices are safe from Meltdown and Spectre. It turns out that the basic lack of caching and speculation in Raspberry Pi is enough to keep things good. If you’re interested, they have an extensive walk-through of the internal logic which Raspberry Pi uses instead of these two techniques. As they conclude,

“Modern processors go to great lengths to preserve the abstraction that they are in-order scalar machines that access memory directly, while in fact using a host of techniques including caching, instruction reordering, and speculation to deliver much higher performance than a simple processor could hope to achieve. Meltdown and Spectre are examples of what happens when we reason about security in the context of that abstraction, and then encounter minor discrepancies between the abstraction and reality.

The lack of speculation in the ARM1176, Cortex-A7, and Cortex-A53 cores used in Raspberry Pi render us immune to attacks of the sort.”

How do we fix this?

For those of us not using Raspbery Pi, the outlook is a little bleak. Meltdown can be fixed with some band-aids and a stronger wall around the kernel. This may cost us all anywhere from 5% to 30% in performance.

Spectre, on the other hand, is a bit trickier to deal with.  It is literally hard wired into the chip. So, we’re looking a long lead time as researchers, engineers, and developers have to come up with a brand new chip architecture or some clever kernel-level work-arounds. How long? Think years.

So, in the meantime, accept all those OS updates coming your way and maybe think about switching to a Raspberry Pi-based computer lifestyle. It’s the only way to be sure.


via JAXenter

January 8, 2018 at 12:14PM

Netatmo’s Smart Home Bot uses AI to let you text commands to your devices

Netatmo’s Smart Home Bot uses AI to let you text commands to your devices


While voice assistants like Amazon’s Alexa and Google’s Assistant are taking over the home, there are still some users who don’t want to talk to their devices. The French company Netatmo, maker of a number of smart home products, wants users to text—not speak—to control their devices no matter where they are. At CES, Netatmo debuted its Smart Home Bot, a digital assistant of sorts that lives within Facebook Messenger that users can text commands to, thereby controlling their smart home devices.

The foundation for the Smart Home Bot comes from Netatmo’s new “with Netatmo” program. Currently, Netatmo devices are compatible with various virtual assistants including Alexa, Siri, and Google Assistant. But Netatmo’s new program will encourage other companies to partner with Netatmo to make devices that work with the company’s software as well as the Smart Home Bot. At CES, Netatmo is showing off a few of the newest “with Netatmo” devices, including smart lights, blinds, and radiators that will debut in 2018.

Any of the “with Netatmo” devices, as well as Netatmo’s own products, can be controlled through the Smart Home Bot via Facebook Messenger. Essentially, it’s a contact that uses artificial intelligence algorithms and natural language processing to decipher text commands you send it to control different smart home devices. You could text the bot, “Who is at home?” when you’re out, and it’ll reply with photos of the individuals that the Netatmo Welcome recognized in your home. You could also text a command to set your home’s temperature to 70 degrees and a “with Netatmo” smart thermostat would be set to your liking shortly thereafter.

Netatmo also claims that the Smart Home Bot will eventually be able to control multiple devices at once, depending on the command. For example, texting, “I’m leaving,” could close the blinds, turn off the lights, and reduce the thermostat’s temperature all in one go. Netatmo didn’t explain how this will work, particularly whether users will have to create a preset for such a situation and assign it a command or if the bot learns your habits over time.

via Ars Technica

January 8, 2018 at 02:02AM

iDevices® Introduces the Instinct™, a Smart Switch that Embeds Amazon Alexa into the Walls of Homes, at 2018 CES

iDevices® Introduces the Instinct™, a Smart Switch that Embeds Amazon Alexa into the Walls of Homes, at 2018 CES

iDevices® Introduces the Instinct™, a Smart Switch that Embeds Amazon Alexa into the Walls of Homes, at 2018 CES



LAS VEGAS – Jan. 7, 2018 – iDevices, the preeminent brand in the connected home industry, is integrating voice service capabilities directly into the walls of any room. When a standard light switch is replaced with the Instinct, users gain the functionality of a smart light switch, standalone voice assistant, ambient light sensor and motion detector — all in one product.

“The iDevices Instinct is a breakthrough in the smart home industry. It represents the next evolution in our extensive line of premium connected solutions,” said Chris Allen, president of iDevices. “The Instinct is one of many innovations to come from iDevices, forever changing the home as we know it.”

With the Instinct, users can not only manage connected products, but also listen to music, find a new recipe and get the news with premium sound by Soen® Audio. Motion and light sensors allow users to automate lights and receive custom alerts, providing energy savings and peace of mind.
The Instinct works in concert with iDevices’ extensive line of smart home solutions and other Alexa-enabled products like the Amazon Echo. The Instinct’s lighting functions can also be controlled with the iDevices Connected app, Siri® and the Google Assistant.

With over-the-air updates, the Instinct is futureproof, meaning users will always have the latest technology installed throughout their home.
The intentionally traditional design of the Instinct fits into the décor of any room, and its voice-activated LED ring lights up to reveal its hidden intelligence.

As with all iDevices products, the Instinct is hub-free and compatible with iOS and Android.
The iDevices Instinct will be available in 2018.

Visit iDevices at the 2018 Consumer Electronics Show

The Instinct will be on display at the CES Sands Expo, Level 2, Halls A-D, Booth #42514.

About iDevices, LLC

iDevices®, the preeminent brand in the smart home industry, is making IoT accessible to everyone with their premium line of Wi-Fi® and Bluetooth®-enabled products. With a comprehensive set of both plug-and-play and in-wall solutions, iDevices manufactures connected power, lighting, and climate control solutions for homeowners, professional installers, and builders alike. Their world-class team of in-house engineers and software developers maintain a forward-looking approach to in-field upgrades and platform integration; ensuring their products always remain at the forefront of home automation. iDevices seamlessly connects people to their worlds with sophisticated technology that enhances everyday life. For more information, visit
iDevices Media Contact:
Meghan Petchel
305-374-4404 x 162

Please enable JavaScript to view the comments powered by Disqus.

via iDevices

January 7, 2018 at 08:36PM