Episode One Hundred and Ninety One: When Things Listen; An Unproductive Argument; Process; Until It Turns Into Writing
0.0 Sitrep
On the way back to Portland from Louisville (Lou-uh-ville), Kentucky after having spoken on a panel on "syndicating innovation across jurisdictions in government". My first conference in the government space (I'm not counting attending the Code for America summit last year, that was more of a throwing in at the deep end before the new job). From my point of view, there were some reasonably interesting parts (and certainly interesting people to meet and have met) but quite a lot of the talk was, well, talk. For example, a lot of talk of the potential role for the US Federal Government in setting standards when my own inclination is not to wait for someone else to do it for you (nor to wait and try to do it 'perfectly') but to start bottom up and iterate as you gon. In other words: just start.
1.0 When Things Listen
The latest realisation that threatened to bubble over into everyday life was the realisation - through the implementation of a ham-fisted privacy policy - that Samsung's Smart TVs that feature voice recognition features, well, listen to you.
Amber Case has written well on this subject[1], explaining that Samsung relies upon a third party provider - Nuance - for voice (and more accurately, speech) recognition capability (and if memory serves me correctly, Nuance also powers the speech recognition portion of Siri's architecture for Apple) to provide the interpreted data it needs to run commands when you use the voice recognition feature. Which makes sense, and is how a lot of these technologies work unless you own a lot of the stack, in the case of a company like Google.
To me, part of this comes back to a failure on Samsung's part to recognise their users' literacy of the systems that go into the operation of the products they sell. It used to be that when you bought a thing from someone, that someone was responsible for all of the things that were in it - and that whilst physical components might come from multiple suppliers, those components didn't go off and tattle-tale. Or, rather, that the product was bounded in terms of connectivity: I can see the edges, and nothing, goes outside of those edges. Smart TVs and smart/connected anything else started to upend that model of understanding, but only to the extent that it makes sense for certain user-focussed features. It makes sense that for a Smart TV to be in communication with a third party like Hulu because, well, you're watching a TV show or a film *from* Hulu. What is unclear - precisely because it isn't shown in a visual manner, for example - is what is being sent *back* to Hulu. Televisions (and especially the formerly single-purpose items of consumer electronics used to receive signals over the air, through cable or satellite) have for most people only ever acted as *receivers* and not transmitters.
I'd suggest that another reason why this revelation that to perform commands by voice Samsung's TVs necessarily need to transmit information to a third party is that, again, these are televisions. As others have pointed out, no-one seems to be particularly surprised that devices like smartphones can (and in some cases do) passively listen at lal times, waiting to be activated by a keyword. And how else could that keyword activation work, if not by passive listening at all times?
But again, we're not used to having such objects in the world. The way they work is not explained to us. And I was thinking the other day, we're taught in schools - at least, I and my classmates were, a good twenty years ago - about biological systems and cycles, but certainly not information systems and cycles. It's not as if we're not capable of understanding flows. But when some of us talk about understanding the world in the form of systems and infrastructure and how those systems interact, the surprise at TVs talking to other things is just yet another example of mental models not having caught up to the pace at which the modern world is being delivered and experienced.
So there's the challenge: in devices formerly known as dumb, where there had been no expectation of *transmission* of information from the device, how do you show, unambiguously, that such transmission has occurred and is occurring?
[1] Samsung TVs and Privacy Policy Gone Wrong - Amber Case
2.0 An Unproductive Argument
If I were a mean writer looking for an unproductive argument and one that would attempt to troll to generate pageviews and an influx of People Expressing Opinions On The Internet, I'd probably start an unproductive argument like Which Company Do You Think Has Less Empathy: Google Or Samsung, And No, You Can't Choose Facebook.
The latest salvo in "do a dumb thing that you can't reasonably expect would be well-received by users who paid for your product, so the only reasonable expectation is that in fact you do not give a damn how your users are treated" is a Samsung entry: the news that they might be inserting ads when you use third party apps on their Smart TVs[1].
At this point, I don't really feel like I need to bang the empathy drum but to say: in what world does a manufacturer think it's a good idea to *surprise* people with extra advertising in a product that they *own* and again, have a reasonable expectation as to how it might function. To paraphrase the great scientist Dr. Ian Malcolm again, I can't help but imagine that the product managers at Samsung behind this particular "feature" - and I must say, as a consumer I am always looking for ways to encounter advertising that is relevant to my interests - were too busy congratulating themselves about having found a way to increase revenue and, well, to deliver advertising relevant to my interests, that they didn't stop to think about whether they *should*.
Score: Samsung 1, Straw-Man Google, 0.
[1] Samsung smart TVs inserting ads into third-party apps - Ars Technica
3.0 Process
That Russell Davies has, in his inimitable and economical way, merely by *quoting someone else* demonstrated that he is yet again a ridiculously smart individual. This time, a quote from Charlie Stross Bond/Spy/Cthulu mashup The Rhesus Chart, from the Laundry Files series:
"A bureaucracy is all about standardization, so that necessary tasks can be accomplished regardless of the abilities of the human resources assigned to it."
When I took my team to visit GDS at the end of last year, one of the things we came away with was being beaten over the head with the concept that if you're looking to see if transformation is going to take hold in an organisation, the two things you want to check for are: research (ie: measurement) and iteration. That's all. Those two things should in theory allow you to fix all other things, given time.
But this quote about bureacracy and standardisation has hit that home again as I've been thinking with our team about how we help transform and prepare governments in America to properly operate in the 21st century which, as weird and complicated as it is, is only going to get weirder and more complicated, and at an increasing rate at that.
Part of this is the increasing belief after having had the opportunity to meet with and listen to people like Sir Ken Robinson[2] - you know, the TED talk about schools killing creativity - is that the majority of people in our workforce have been churned out in a more or less industrialised fashion with skills and training pretty much still fit for an assembly line. Again, I completely acknowledge that I'm spouting off with my *reckons* and that the last time I really had direct experience with k-through-12 education was, well, when I was *in* it, and the occasional chat with teachers every now and then, but the anecdata is that we wouldn't still be talking about a purported "skills gap" if there wasn't, well, a purported skills gap.
To bring this back: government is great at bureacracy and arguably it was the British government that hit upon it as a great idea (that may well have been a perfectly reasonable operational response at the time). But as the world is coming to realise, we just don't live in a world where a bureacracy with standardisation - and the silos that come with it - is fit for purpose.
So with that in mind, or in the back of my mind at least, we've been talking more at work about the end-to-end service delivery team, which isn't as snappy as a Multi-Disciplinary Team, but is at least more understandable to American audiences and makes more sense to them outside the domain of the delivery of digital services. What we've found when we've been talking to governments has been that when we mention end-to-end service delivery teams as an explicit unit we're interest in increasing the capability of, their eyes light up in a way that indicate that they'd really like to invite us in and help demolish some organisational silos. Which is always a nice thing to see, because who doesn't like to wake up in the morning to the smell of demolishing silos?
In other words, this is a recognition that the necessary tasks in the delivery of a service are specialised and require unique roles within a team that works together. Which if I were Russell, I'd have found a better way of saying. In other words, the unit of delivery is the team.
[1]blog all kindle highlights: The Rhesus Chart - Russell Davies
[2] Ken Robinson: How schools kill creativity
4.0 Until It Turns Into Writing
Two last things. First, Matt Webb in the latest of his blog posts wrote:
But let's be clear... this is all about me: What I get out of this is that somehow, by typing, four unrelated things that have caught my eye sometimes show signs of coherence. I get glimpses of the gestalt. So that's why I type.[1]
which is pretty why *I've* been writing this newsletter, write back in the beginning. This is a way for me to get my thoughts out and to see what they are. I've found that I do a lot of my thinking out loud, or verbally, or at least in sentences and that it's only when I'm writing or talking to ideas start to take shape. I'm not a doodler or a sketcher in the way that some of my friends are (which I'm trying to take not necessarily as a deficiency, but just in a way that I'm different; regardless, it's always been on my list to learn how to draw better) but I do know that when I write - and readers of this newsletter who've followed my progress over the past year will have noticed this - you can pretty much see my thoughts and arguments developing over the course of paragraphs if not even the very sentence that is running on.
And, of course: the death of David Carr, whose writing I've always admired, and for expressing the other reason why I started the practice of writing every day. I wanted to keep typing until it turned into writing.
[1] Filtered for Muybridge and Moorcock - Matt Webb
--
OK, that's it. Friday, 7:58am Eastern and I'm signing off until the next time. This was fun. Let's do it again. And if you're new, say hi (and check out the archives), and if you're old, say hi too, I'm sorry I haven't been around as much. Perhaps it's given you a chance to catch up on what I hear is in a number of cases a large folder of unread newsletters. And in any event, send me notes. I love receiving them.
Best,
Dan.