Episode Nineteen - Not Trying Is A Signal, Peak Game, EasyHard, SnapChat
1. Not Trying Is A Signal
I wrote yesterday, in my bit about Fax Your GP (which, maybe not on its own, but certainly appears to have contributed to this recent decision to push back rollout of the care.data database a little[1]), that with ever increasing emphasis on customer service and user experience, the delta between what's good and what's intolerable inexorably decreases. That's to say: once you've seen something with a good user experience, it's hard to justify other experiences in the same category having a shitty user experience.
Sometimes, this makes sense: you can pay a little more and pay Virgin America and get a super-good user experience when you fly, compared to when you fly United and you're wondering why they're not paying you instead. That can kind of make sense when you're flying because you don't really have that many options.
But in other cases, where the goods or services are highly substitutable, the distinction between one option which has a pleasurable service (especially one rendered digitally) and one that isn't is just going to lead to instances of nerd rage. A slightly more mellow than nerd rage case in point, Russell Davies (again) on Sony's new product[2] and the fact that while they're entirely capable of a singularly impressive engineering feat, everything apart from shipping the product and making it falls down from a new customer's point of view. And that's despite the fun stuff[3] you can do with it.
And all of this in spite of the fact that you just know there's someone in management, somewhere, saying that they know a fifteen year old kid (hopefully, equally likely to be a girl than a boy) who could 'knock up a website' that would do that in a few days.
Nick Sweeney helped me articulate this, in the context of government, a little better in a series of emails:
* If government can't produce a good, digital user experience, when other entities can, then government looks bad (see: Healthcare.gov)
* If government *can* produce a good, digital user experience (see: the work of GDS in the UK), and then for some reason a good digital user experience isn't produced (see: care.data opt-out), then suspicion as to failure of implementation includes policy reasons (ie malice) as well as incompetence.
So: companies and governments. You're on notice. It isn't hard to do this kind of thing. It isn't easy, either. It's just simply doable. The fact that you're not doing it is now a valid signal that you're not doing it for a reason.
[1] http://www.bbc.co.uk/news/health-26239532
[2] http://russelldavies.typepad.com/planning/2013/10/death-to-innovation.html
[3] http://www.separate--together.com <-- disclosure: work thing
2. Peak Game
Just a few minutes ago I learned that Irrational Studios, of System Shock 2 and Bioshock fame, was effectively closing down and being rebooted as a leaner 15-person-plus-Ken-Levine studio[1]. The backdrop to this happening is interesting: I quipped on Twitter that one way of looking at this is in the context of the eternal war between the Triple A expensive 3D console tentpoles (your Bruckheimer/Bay equivalents) and smaller shops that produce work a bit like Telltale Games do with their critically acclaimed (and, by all accounts, commercially successful) Walking Dead series. What Levine is interested in sounds, well, interesting:
"I've been working on a concept I call narrative LEGOs," he said, "which is how do you take narrative and break it down. What are the smallest part of narrative that you can then remix and build something out of? Mix and match."
What's even more striking is the comparison that many made when Bioshock Infinite came out to a game almost on the opposite side of the spectrum (and with many even taking issue with whether it could be called a game in the first place): Portland, OR based Fullbright Company's Gone Home. And last night, watching a friend play The Stanley Parable (which game's narration instantly reminded me of The Guide in The Hitch-Hiker's Guide To The Galaxy).
Everything, they say, is being disrupted and eaten by code. But some of that disruption is more slight and, well, I would say insidious, but let's just say longer lasting and less obvious. Videogames, out of all media, were bizarrely one of the arenas of culture that moved the slowest to embrace digital distribution, with the major publishers clinging on to shipping boxed product. While the combination of digital distribution and massively available pocket consoles in the guise of phones made attacking the low-end easy, it feels like it's only until recently have people started to cotton on to the fact that you can make Big, Meaningful, Story-based experiences to a reasonable (read: profitable, I think) audience.
So: here's the prediction, and it's not a very insightful one, I'm afraid. I think we've hit Peak Expensive 3D FPS AAA Story Game, modulo things like Michael Abrash's dream of VR Within Five Years Yes Really I Mean It This Time.
My money's on Levine experimenting with mobile and shipping fast.
As a parting aside in this section: a metric crap-tonne of Measurable Attention Time used to be taken up by daytime TV and soaps. Those "video games" should see that as a stupendous opportunity, but the business model behind it might need a bit of tweaking.
3. EasyHard
Via the inestimable Matt Jones, Scott Jensen on EasyHard problems[1], particularly in the realm of home automation, which inevitably touches upon the internet of things and deciphering the intent of us meat-based-water-bags.
Jensen takes some time to remind us of Moravec's paradox: that it's bizarrely "easy" for us to make computers do what might appear to be intelligent tasks but are in fact ideally suited to being computationally brute forced by just walking, rapidly, and with purpose, down every branch of every tree of a possibility space - e.g. solving Chess. Conversely, it's super hard (read: we haven't really done it yet) for apparently easy tasks that we take for granted like, oh, I don't know recognising cats (which takes a cluster at one of the largest private computing grids in the world), or walking or spotting things or even understanding what it is that I'm saying.
A lot of Jensen's examples are of the "Do What I Mean, Not What I Say" problem, which is probably part of the fault of the promises of perceptual and ubiquitous computing coming home to roost. Until you actually sit down and deconstruct all the dialogue that the actors have in conversation with the Computer on Star Trek: TNG (which is a pretty good Tumblr project, to be honest) and parse all the meaning out and how phenomenally intelligent Majel Barret is, it's hard to have a proper appreciation for the sheer amount of contextual awareness and cues that a home automation system needs to have. In a sense, some people have been saying this about the nature of artificial intelligence itself: that it's hard (or may even be impossible) to create the new bar of what we raise to be artificial general intelligence without creating one that's embodied. And the same may be true for home automation or other types of intelligence.
Part of this, when we're still in the canyon of EasyHard is to have easy to understand, transparent systems where you can understand why something has happened, and how to either make it happen again or not. All of this is to say that repeatable expectable behaviour is one of the things that I'm assuming without having to go look for a citation is the kind of thing that reduces stress in humans, as opposed to unpredictable random outcomes when you want your lights turned on.
[1] http://jenson.org/easyhard/
4. Snapchat
Here's a super short one. Why are we surprised that a generation of people who have been told, in no uncertain terms, "Be careful what you post on the internet! It will live forever and haunt you!" have flocked, en masse, to ephemeral conversation tools like Snapchat?
Bonus thought: Phil Gyford's insight[1] that perhaps the nicest thing about SnapChat is its complete lack of archiving. Us old people have grown up with everything disappearing and upon the availability of digital media instinctively wanted to Save All The Things. But perhaps a more mature attitude is that some things can be saved and some things - well, not so much. And I can see that not having to see the entire history of a conversation is freeing in a very interesting way.
[1] http://www.gyford.com/phil/writing/2014/02/17/snapchat.php
--
It was good to get back into the saddle again today. I have a bunch of things I want to talk about tomorrow, too, it sounds like.
As ever, blah blah about all of your feedback, but only substituting words that convey a sense of genuine gratitude at the feedback that I do receive.
Dan