The Annoyance of Today’s Personal Computing – Amazon Echo Dot

Before escaping the dreariness of a Seattle winter for a few days, I ordered some holiday presents, so we’d have something to play with on our return.  Having just watched the first Iron Man recently, I was curious as to the state of the ‘Jarvis’ form of personal computing.  Then I read the article by Mark Zuckerberg and I thought, well, alright then, I’ve got to give this a go.

Full disclaimer as usual; I’m a Microsoft employee and have been for 18 years, but I’m not company apologist.

So, I placed the order and off we went.  Upon arriving home, there was a package (not stolen from the porch oddly enough), and I thought, “what might that be?”.  It was tiny.  I opened it up, and sure enough, it was the Amazon Echo Dot.


I purchased it on a sale of $40.  It’s since gone up by $10.  I was really surprised how small it is.  It fits in the palm of my hand roughly.  At this price point, it’s almost an impulse buy for a tech enthusiast.  Roughly the same price as a functional Raspberry Pi, even cheaper considering the Pi need a power supply, uSD card, monitor, keyboard and mouse.

Setup was a relative breeze, with the help of an app, which required the iPad, or other non-Windows tablet, to setup.  Press and hold a button for about 5 seconds, tell it your home network’s wifi password, and you’re all set.

Then you start talking.  Right out of the box, the easiest things to test are playing of music and asking about the weather:  Alexa, what’s the weather like?  Of course, we just came in from outside, so we already know, but it was fun.

My wife got into it by asking to play some Hindi songs: Alexa, play hindi songs.

Alexa obliged, and started playing from some random play list of hindi songs.  Then we tried to get specific:  Alexa, play Chammak Challo…

That one took awhile with several miscues, and offering up things we’d never heard of.  It did finally figure it out, but I don’t know if we’ll have to go through a whole series of phrases to get it to play again.

After doing the whole: Alexa, play… for a while, our 3 year old figured out what was going on.  He’s starts shouting “Alexa, play…”.  For some reason, Alexa doesn’t respond to his implorings.  I don’t know if that’s a blessing or bad design.  I’m sure there’s a way to train it to recognize his voice.  I’ll wait and see if I want to give it that particular skill.

The echo system comes with things called ‘skills’.  These are various phrases tied to various actions.  There’s a whole marketplace of free skills.  So, you can do something like “Alexa, tell Uber to order me a car”, or whatever the phraseology is.  You ‘tell’ another skill to do something on your behalf, as if each skill was it’s own AI.  Meh.

What I like about it is the ease with which you can quickly integrate this into your home life.  We stuck it in the kitchen, which is central to our daily activities.  I also purchased a couple of wifi controllable lights, so that we’d have those to play with.  I have the intention of installing them in one of the rooms, perhaps downstairs, just so we can say things like “Alexa, set some action lighting in the play room”.  That and some other home automation might be interesting.

The annoyances…

Sitting around the kitchen, saying “Alexa” all the time is really annoying.  I not a natural flow of conversation to say a thing’s name every time you want to address it.  Try this at home, every time you want to say something to someone, say their name first, even if they’re standing right next to you and you’ve been talking to them for the past hour.  It’s really annoying.

We yell at Alexa.  Not quite knowing what volume is best, we shout “Alexa” all the time.

We want a different wake word.  I don’t think this is possible, maybe you have a couple of choices.  But, I’d much rather say “dufus” than alexa.  And of course I’d like to get the cool Jarvis voice, or Majel Roddenberry, or a selection of voice actors to interact with.

So far, the major annoyance is the constant talking, and the inability to do simple things like ‘save this song to my favorites’.  The latter doesn’t necessarily work unless you have Amazon Music, or whatever.  Some cryptic message, and lack of instructions on what to do right away to save the song I’m listening to is just annoying.  I just wanted to say “Alexa, I like this”, as the instructions said, but couldn’t quite make it happen.

Of course this is an Amazon product, so there is purchasing involved.  It’s tied to my Amazon account, and I can order a $20 loaf of bread easily enough.  I’m sure this particular feature (shopping) is well tuned, but I haven’t tried it out yet.

The other aspect that’s interesting is the license.  I actually bothered to read it to see just what kind of shenanigans they were up to.  Of course, saving your utterances is in the mix.  Saving your orders, music choices, preferences, all of that stuff.  And passing it along to the appropriate partners at the appropriate times is fully available to Amazon.  What strikes me is how easily I’m willing to give away my personal space to a company.  Not only am I dropping a bugging device into a central location in my home (besides the cell phone I carry with me at all times), I’m giving them the right to save everything I say and use it against me in targeted advertising.  At least I turned off the notifications portion of the app, so hopefully Alexa won’t be whispering subliminal advertisement as I eat my morning cereal.

The other aspect is how easily I give up control of my home.  I’m willing to let a cloud attached device sit on my home network (what an effective spy tool), and talk to my lights, security system, garage door, HVAC…  I’m setup, even better than the nest, to get my home hacked from abroad with ease.  I can imagine the scenario.  Alexa knows when everyone’s out of the house for extended periods of time, either through observation, or because I’m actually using the calendar function.  Alexa/Amazon knows everything I’ve purchased in the last year.  Alexa controls the electronic locks on my doors as well as garage opener.  The hacker sitting in a non-US expeditable jurisdiction hacks the echo dot, passes information along to a local actor, and opens up the doors at the appropriate time for my home to be swept clean.

This scenario will surely play out with the appropriate apologies from Amazon, or whomever first gets hacked in this scenario.

Why I’m not worried.  Although I have a security system, I won’t be hooking it up to Alexa, or any other cloud attached thing that I don’t control.  I won’t give Alexa access to the garage, and I prefer good old fashioned lock and key.  But, this brings up another point, and kind of the crux of the state of personal computing.

Who’s data is it anyway?

In this modern cloud rage fueled economy, the likes of Google and Amazon have claimed our personal data for themselves and their own purposes.
The Echo Dot is simply the ultimate in personal data collection devices.  Never before has a company (besides Nielsen) had a device in your home collecting so much, and possibly controlling it as well.  This is where the personal computing thing breaks down right now.  If it’s going to be personal, I think I need to be in control of the data.  I can believe this whole voice activated thing needs to be cloud based for the moment, because you can leverage the collective phrases of the entire planet to figure out what’s being uttered, and what action to be taken.  But, come one, give the hardware another couple of years and you’ve got to believe that the device itself will be strong enough to do that all on it’s own.  I’ve got a few terabytes of storage, and more MIPS sitting idle that an entire data center of a small company.  Surely I’ve got enough capability to figure out “dufus, play ‘what is love'”.

So, a single Echo Dot has made its way into our home.  It performs about as well as I expected.  The audio coming out of the device itself is actually better than I expected from the reviews.  It’s not Jarvis level personal computing, but I can certainly see the makings.

Microsoft has a ton of voice recognition, and Cortana, and ‘skills’ bot stuff as well.  Microsoft does not have a $40 piece of hardware to pull it all together.  After looking at the drool worthy Surface Studio announcement a couple months back, I’m hopeful that my company can come up with a simple device to pull all our tech craft together and play in this space.  It’s going to be huuugee, and I don’t want to miss it.

My data is my own, and the first company that fully embraces that perspective gets my money, and respect, big time!


Note To Self – VS Code seems reasonable

No secret, I still work for Microsoft…

Over the past 17 years of working for the company, my go-to editor had largely been Visual Studio.  Since about 2000, it was Visual C#.  Then around 2011, I switched up, and started doing a lot of Javascript, Lua, and other languages, and my editor went from Notepad++, to a combination of Sublime Text and vim.

Most recently, I’ve had the opportunity to try and enable some editing on Windows 10 tablets, and I chose a new editor, Visual Studio Code.  I am by no means a corporate apologist, but I will certainly point out when I think my company is doing something good.  Visual Studio Code is an easy replacement for Sublime Text, at least for my needs and tastes.  I’ve been trying it out on and off for the past few months, and it just keeps improving.

Like all modern editors, it has an ‘add-on’ capability, which has a huge community of add-on builders adding on stuff.  Of course, there’s some lua syntax highlighting, which makes it A number one in my book already.  But, there are other built in features that I like as well.  It has a simple and sane integration with git repositories right out of the box.  So, I just open up my favorite projects, start editing, and it shows which files are out of sync.  A couple of clicks, type in my credentials, and the sync/push happens.  I’m sure there’s an extension for that in all modern editors, including Sublime Text, but here it’s just built into the base editor.

One item that struck me as a pleasant surprise the other day was built in support for markdown language.  I was refreshing the documentation files for schedlua, and I was putting in code block indicators (“`lua).  After I put in one such indicator, I noticed the quoted code suddenly had the lua syntax highlighting!  Yah, well, ok, getting excited about not much.  But, I had never seen that with Sublime Text, so it was new for me.  That was one of those features that just made me go ‘sold’.

The editor has other features such as being able to run a command line from within the editor and such, but it’s not a full blown IDE like Visual Studio, which is good because the tablets I’m running it on don’t have 4 – 8Gb of RAM to run Visual Studio comfortably.  So, it’s just enough editor to replace the likes of Sublime Text.  I also like the fact that it’s backed by a large company that is dedicated to continue to improve it over time with regular updates.  The community that’s being built up around add-ons seems fairly robust, which is also another good sign.  Given Microsoft’s current penchant for Open Sourcing things, I would not be surprised if it showed up available on GitHub some day in the future, which would just make it that much more interesting.

So, for now (future self), I will be using VS Code as my editor on Windows, MacOS, and Linux.  It has the stability and feature set that I need, and it continues to evolve, adding more stability and features that I find to be useful.