s07e24: What Kind of Day Has It Been
0.0 Sitrep
Saturday December 7 2019.
I’ve been trying to write this episode since Monday December 2, which was a good few days ago. I was in Sacramento for work Monday through Wednesday, and writing the newsletter was hard the week before too, I think because coming back after taking a couple of weeks off after Thanksgiving was difficult. I was in Sacramento for work Monday through Wednesday, and writing the newsletter was hard the week before too, I think because coming back after taking a couple of weeks off after Thanksgiving was difficult.
I tried again on Wednesday December 4, with some time to write on the 90 minute plane ride home, and was stymied because the wifi wasn’t working on the flight. ecember 4, with some time to write on the 90 minute plane ride home, and was stymied because the wifi wasn’t working on the flight.
Here’s the introduction I wrote, then: ere’s the introduction I wrote, then:
0.1 Wednesday December 4.1 Wednesday December 4W
Wednesday December 4, 2019. Here’s a typical “oooh, where does all the time go, hm?” - it’s already December and people keep pointing out that it’s nearly the end of the “decade” or something, which means we have to critically assess our personal achievements over the past ten years, as if that’s a thing that defines our worth.
Anyway:
I’m writing this on the way back from the day job in Sacramento, but the wifi isn’t working on the plane. To be honest, I’m not that bothered about the wifi not working because a couple days ago I’d been reminded about the time a fan blade had catastrophically failed (i.e.: broken) on a Southwest 737NG jet, smashed through the engine housing/cowling and broken a passenger window, resulting in the passenger at the window seat being partially sucked out and suffering injuries “incompatible with life”. There is a point to this anecdote / “thing that caught my attention” and it’s not just a strong personal re-assessment as to my window-seat preference.
It’s that I feel like I’m seeing a bunch of “the world is too complicated and we’re building things that are just past some sort of criticality point that maybe we’re not quite ready for” articles. I feel like we had a bunch of these that time the Icelandic volcano Eyjafjallajokull went off and disrupted the global supply chain. Or that time Maersk got hit by the NotPetya malware and again the global supply chain broke and people were worried about getting, I don’t know, anything, because who really knows the dependencies these days? The most recent example was I think a Crain’s Chicago Business article about high-efficiency jet engine technology used in the latest passenger jets. We’re not going faster, but we’re being more efficient with fuel. From what I understand, that means we’re making bigger engines (part of the issue with the 787MAX8 is the fuel-efficient maeengines needing to be mountmaeed imaen a different place on the wing because they’re bigger), and bigger engines means bigger fan blades and bigger fan blades means higher temperatures and new materials science to make sure (I dunno) the blades can go Really Fast And Really Hot Without Exploding.
I mean, in general this is good! More efficient air travel is good! It turns out that also making these fan blades is hard and maybe we shouldn’t be using them yet? The manufacturers appear to be having problems with them, but we’re still using the engines anyway because there’s some sort of deep imperative (I’m looking at you, late-stage capitalism) to make sure that airlines keep growing and stay profitable and that more people fly even though there’s pretty good evidence that all this flying isn’t that great for our continued somewhat-comfortable existence on this planet.
As ever, I am going to turn this into a parable about computers. My friend Rachel Coldicutt has a wonderful phrase that’s now on a t-shirt of “Just enough Internet” which I like for many reasons, not least of which because it’s a wonderful play against Daniel Ortberg’s infamous “What if phones but too much” description of the Black Mirror series.
We keep putting Internet in things and I am not quite sure why. There is certainly some utility to some Internet in things, but sometimes we put too much Internet in things, or the kind of Internet we put into things ends up being too much, just by accident. A recent example that stuck in my head was the fact that high-speed wifi relies on something called MIMO - having multiple antennae that send and receive at the same time because if you do lots of complicated maths really quickly, you can use all of the antennae to send and receive lots of information at the same time. It just so happens though that you can do roughly the same maths to figure out if there are Things Moving, and hey, those might be Humans and before you know it, your home wifi router is telling whoever it wants to whether there are people in your house, whether they’re moving, how often they move and, well, anything else you want.
It feels, still, that we are ill-equipped as a society to understand quite how stupendously the act of lots of maths and processing and networked computers changes how we think the world works. I mean, there’s a model of the world in our head that goes a bit like “well, it’s probably not a good idea for police to install microphones in all public spaces and record everything”, and yet this is pretty much exactly what’s happening once law enforcement in the U.S. realized that it could use social media monitoring tools to look for persons of interest. I mean, it’s not like there was literally a TV show about this, but maybe the problem was that was a TV show that was a drama (even though it was on CBS and reviewers could use phrases like “ripped from the headlines”), but maybe it’s because it wasn’t a documentary narrated by Morgan Freeman?
My (repeated, tired) point is this: things are very complicated and we appear not to have yet produced tools that help us understand and navigate the complexities of that world, that help surface the consequences of that world. I feel like it’s one thing to say that we live in an Information Age or an Information Economy but honestly it feels more like we live in 99% of a Electronic Filing Cabinet Economy and 1% of an Information and Data-Informed Decision Making Economy.
Somewhat related to this, there was what felt like a parody paper published in [] by a [] from Accenture pointing out all the problems with Human Decision Making, and that maybe what with all the flaws in Human Decision Making (inability to explain decisions, built-in bias, etc) we should hold off on implementing Human Decision Making in things until we get a good idea about what’s going on.
I’m not sure if this was intended to poke fun and be part of an argument for machine learning-backed decision making (I mean, it sure felt like it), but most people I’ve talked to who’re in the “hang on a second” camp on machine learning and algorithmic decision making are *also* and frequently exactly the same people who think that human decision making also needs to be improved! It’s this bizarre straw man argument where the implication is that if you’re against machines making decisions (although, you know, at our direction and on our behalf), then you’re automatically pro human decision making. Like, not really? They can both be not good enough and we can still deserve better from both?
I’m not entirely sure where I’m going with all of the above, to be honest. Just that it feels weird to be optimistic about technology, and part of the reason that I’m optimistic about technology is that I’m by default optimistic about humans. What I’m not optimistic about is blindly blundering through some sort of cavern of technological choices where we just do things because of a sort of hold-my-beer mentality.
But!
This position (maybe slow down a bit and be more deliberate), where one argues for better understanding user needs and then deliberately inventing technology to meet those needs is opposed by the view that invention just happens anyway. New technologies come first because people are literally hold-my-beer machines and can’t stop themselves from being curious. We are the monkeys who have been taught to be curious and to make tools and then think later about what we might use them for. Necessity is not then the mother of invention. Invention happens and then creates necessity. (Those of you with STS backgrounds may have figured out that I’ve recently been exposed to Kranzberg’s laws for the first time).
I suppose I can close the loop on this by saying: now I’m looking at the engine out the window differently.
1.0 Some things that caught my attention
1.1 … of which some medium-sized things
PowerPoint
There’s a PowerPoint Game Jam! For those who don’t know, PowerPoint is an app from Microsoft that lets you make slide shows and put lots of text against bullets on a screen. A Game Jam is a community challenge where people get together and make games (art) under certain constraints, one of which is normally something like making a game over a weekend. PowerPoint is usually used to make boring things and games are not boring, so this thing should be interesting. It reminds me of the roguelike I’d been writing earlier this year set in a dystopic tech company.
Terribleness
I really liked this phrase about the American Automobile Association, describing them as “basically a lobbying shop for terribleness masquerading as a roadside assistance club” in a thread from Waldo Jaquith about, well, how terrible the AAA is and how it doesn’t do things like offer charging services for stranded electric vehicles. I am now on the lookout for more accurate descriptions that fit the “basically a lobbying shop for terribleness masquerading as x” format.
This Product Certified by Mozilla
I was not aware that the Mozilla Foundation has a privacy standard through which it reviews devices (here’s its review for the Facebook Portal) but to be honest I’m not entirely sure people know what a) Mozilla is, or b) what it’s for. I bet a slightly larger number of people know what Firefox is, but even that number’s not got to be very big these days. Anyway, Mozilla describes itself as “a global non-profit dedicated to putting you in control of your online experience and shaping the future of the web for the public good”. At first glance I would say that sentence is too long and that I don’t know who it’s for. Why should anyone care about Mozilla these days? I mean, there are lots of reasons to care about what Mozilla cares about (I think?). This is the kind of commentary you get when someone who’s worked in comms and marketing/advertising looks at a piece of About Text, I suppose.
There’s a part of me that thinks Mozilla has the problem the wrong way around. Yes, people want safe, non-creepy smart speakers, fitness trackers and wireless headphones. But I think they want smart speakers, fitness trackers and wireless headphones *more* than they want safe ones? I’m intrigued to see if there’s any research that says a significant number of people (any number that’s worth it?) would check out Mozilla’s guide for smart device gifts instead of, you know, anyone else’s guides.
Anyway. I have friends who work at Mozilla, or people who are at least Mozilla-adjacent and this is to say that I do not think the reviews are a bad thing, just that I would want to understand what their plan is for them to be effective, or how they know if they’re doing a good job.
AI copyright
I saw an interesting article about A.I. copyright considerations. I am mildly interested in this because I used to be a lawyer and intellectual property was supposed to be one of the things that I was good at. But, my knowledge and experience is at least 15 years out of date now! The other reason why I’m interested is because I’ve “experimented with A.I.” for some creative work and I’m reasonably sure that *I* would have copyright in the resulting work, if only because I did the work of creating the prompt text and selecting the final set of work. That is to say, I made a list of A.I. generated British placenames and I would like to think that the collection of placenames vests in me and that, to the extent that the individual placenames are copyrightable, there was a creative element involved in selecting them in the first place. I think?
Deep learning at the end of the universe
I did not have “deep learning comes up with a grand unified theory within the next 15 years” on my bingo card, but the news that Google threw a bunch of TensorFlow at M-Theory (nee string theory, that bit of maths that fell into the 1990s from the future and promised to explain the universe, only accidentally explains a whole bunch of universes, so is rather lacking in predictive power) and... came up with some new M-Theory solutions? There is something colliding in my head here about deep learning being able to think (sorry) about things in a way that might be fundamentally unlike how we think about things (what with our embodied cognition and all), an example of which is the story being bandied around in the press that Lee Sodol has given up on Go after having been beaten by AlphaGo (itself succeeded by AlphaZero). One of the things I remember Sodol saying at the time was that AlphaGo showed him fundamentally, well, non-human ways of playing Go that we just hadn’t thought about as a species. Which is pretty cool! And, kind of loops back to a feeling about Google throwing TensorFlow at stuff to see what sticks.
People in tech have no idea, really
We may think we’re all about the knowledge and information economy, but the U.S. Department of Education came back with their most recent Adult Competencies report and, well, it’s not great! There’s a set of skills called Digital Problem Solving and there hasn’t been any significant improvement since 2012/2014. IN fact, the number of people who have below level 1 skills increased by a percentage point which, to be fair, we could just attribute to error. There’s two ways of looking at this, one of which is: oh my gosh we need to teach people to computer better. This isn’t even about coding. The second one, which I’m much more in favor of, is: oh my gosh computers are horrendous and stupendously difficult to use. There is so much room for improvement! Digital Problem Solving may as well be “what proportion of our adults have stockholm-syndromed themselves into Thinking Like A Badly Designed Computer” in order to figure out how to accomplish a task.
Has anyone considered the negative externalities
The European Emergency Number Association (which is an association for emergency service response in Europe and not, say, an association that will assign you a number if you need one in an emergency, like a sort of URGENT BROOKLYN INTEGERS) is concerned that automated emergency services communications from wearables are massively increasing the number of false-alarm emergency alerts. Their paper pretty much says that “emergency services are often not involved int he development of these features, which can result in flawed communications” and “this lack of adequate communications is a consequence not anticipated by tech companies, which can hinder the work of emergency services.” I mean, this is my favorite thing of Not Considering Or Anticipating External Consequences By The Tech Industry In General! The high level tension here of course is the belief on the one hand that if you had to go around and actually talk to emergency services before you rolled out something like automated emergency services calling if you fell over then you’d never get around to it because everyone would want to say no. Or that it might be very difficult to get the buy-in. Better, then, to just launch the damn thing! Or maybe to talk to one or two emergency services people. I have no idea if Apple, for example, even briefed *any* emergency services responders on what to expect about a feature like automatic 911 calling beyond fulfilling any regulatory requirements. Anyway, here’s EENA’s position paper, which seems reasonable if only because it’s just trying to point out “hey, maybe you could involve us in some way?”
1.2 Even smaller things
Seriously, this bit is pretty much random association, or a selection of thoughts that join two things together:
There was fleeting reference in one of my myriad miasma feeds of a new car that had three in-dashboard giant displays, and my thought was: what does it signify when the first car ships with a high frame rate HDR display? I mean, to me it means that either a passenger’s going to be watching a Prestige Television Show that’s been commissioned by, I don’t know, Tesla, who will be spending billions of dollars on content for their in-car entertainment network (DON’T LAUGH, YOU KNOW IT MIGHT HAPPEN), or we’ve got usable/lawyer-friendly autonomous vehicles.
I liked the short story, The Time Invariance of Snow as a re-telling of The Snow Queen by E. Lily Yu
Someone on Twitter was setting up their new Pixel phone and there’s a bit of copy for face unlock onboarding that says: “How to set up face unlock See below to learn how to rotate your head” which made me think in quick succession: a) really? b) really?! c) Huh, I remember Young Adult novels have a thriving business of consultant sensitivity readers d) I will be your sensitivity reader for seriously weird copy like “learn how to rotate your head” which looks like someone just tried to copy Apple’s Face ID copy (the onboarding screen even comes with a similar face rotation illustration) but in a way where they wouldn’t get in trouble with their high school English teacher for cribbing. I mean, I know how to rotate my head. What you mean is hey, rotate your head like this to set up your phone.
I wrote above about the time Maersk got punted back to the paper age because of Notpetya and it so happens that the Philadelphia Inquirer just published a piece on a Notpetya’s “cyberattack” on Merck in 2017. The piece is an interesting reminder about how insurance companies are reacting to claims related to cyberattacks (either paying out, or disputing that they were cyberattacks and instead excluded due to being “acts of war”), and it is only at this point that I am starting to feel like we’re entering the cyberpunk phase of hostile digital wildlife that continues to live online because, well, we’re as shit at vaccinating our digital infrastructure as our biological infrastructure. Notpetya’s not going to go away because our digital herd immunity is crap, too. I mean, could you imagine what it would take to eradicate a certain digital attack in the way we’ve done polio?
A phrase that came up at work as we were discussing an existing (legacy, etc.) computer system for a government department as one that’s concerned with administering a program and not delivering outcomes (ugh, I dislike that latter phrase) and that the computer system viewed in that light is closer to a filing system than something that, say, makes sure policy is effected as intended.
That is enough for now, I think.
I turned 40 a couple weeks ago and my wife threw a surprise birthday party for me last night which, I have to admit, was a legitimate fucking surprise. I mean, I thought something weird was going on, but I did not think it would be a surprise birthday party dinner, and I did not think it would be quite like the way it turned out. I don’t think I’ve legitimately been surprised like that before and if I’m being completely honest, I’m still a bit emotionally wrung out by it. And the subsequent karaoke.
I hope you’re well. Where I am, the days are short and dark, Christmas lights and decorations are spawning all over house exteriors and there’s that consistent patter of Amazon deliveries against our front door.
See you again next week,
Dan