s09e10: Why Not Put An RFID On Everything, and Griefing
0.0 Context setting
It’s Sunday March 21, 2021. I have been playing a lot of Hades lately.
I am sure there’s nothing to be read into me getting into a game that involves fighting your way out of the underworld only to just die and die and die and die and die over and over again.
Two big things in this episode, and then a couple of smaller things. Let’s go:
1.0 Some things that caught my attention
1.1 Checklist item: put an RFID on everything
Jacob had commented on software and that the consequences on getting it “wrong” are lower than that of Emergency Medicine [tweet], and here was my chain of thought:
- Yes, but perennial example Therac-25
- More recently, from October 2019, the peer-reviewed paper in Nature, Millions of black people affected by racial bias in health-care algorithms
- IBM’s Watson Health potentially underdelivering against its promises [Computerworld], see also IBM’s Watson supercomputer recommended ‘unsafe and incorrect’ cancer treatments, internal documents show in Stat, from 2018.
So, given that software is used in medicine anyway, the risks are interlinked. But I suspect that’s not what Jacob meant, anyway.
But what if there were small bits of software that weren’t involved in, say, diagnosis decisions, treatment plans or even record management? That gets me thinking about
- Atul Gawande’s Checklist Manifesto [Wikipedia]
- … but see also the 2015 Nature feature, Hospital checklists are meant to save lives — so why do they often fail? (Spoilers: badly written checklists, checklists that aren’t appropriate to context, lack of adoption/resistance)
- … but safety checklists in a medical environment are a bit like automated testing in continuous integration and deployment, right? (ha! Simile!)
- Automated testing is generally good, right? I mean, it goes bad when people go for a metric of coverage and you’re in the whole metric-capture situation where suddenly it’s about the metric and not the originally intended outcome, and “test coverage” means people are missing opportunities for other… kinds of testing?
- … so what might automated testing in the context of test before you go to production look like in a medical environment?
A possible answer! Ring, but for objects left inside people after surgery! These objects are called retained surgical foreign bodies [Retained surgical foreign bodies: a comprehensive review of risks and preventive strategies, PubMed; Retained Surgical Foreign Bodies after Surgery, Open Access Macedonian Journal of Medical Sciences].
Look, if I were to have an idea, and foolishly not check as to whether it was a good idea, then one thing you could do would be to install a bunch of high resolution video cameras in operating rooms (and maybe… on people?) and (of course, checking whether they can actually see anything) then run live image recognition on that streaming video to look for anything that shouldn’t be there. My entirely uninformed reckon is that there might you might not even need to do object recognition because, well, artificial foreign bodies tend not to look like living, non-foreign bodies? Current ways of preventing this include checklists (see my train of thought now?) like counting everything, better communication (counting everything out loud, and I guess maybe also involving the practice of saying what you’re doing and pointing at it, and for those technologically inclined, putting an RFID tag on everything and, I guess, hoping that a) you don’t leave anything inside the patient and b) the patient never gets scanned at a library.
Anyway: Ring, but for retained surgical bodies.
PS. I mentioned Therac-25 as one of the seminal cases of Software Killing People, but I realized that the paper finding that an Excel error “led to an estimated 1,500 deaths in the UK during 2020” [Ethan Mollick] due to a conflict between an old Excel file format (.XLS) and the newer one (.XLSX). [Medrxiv paper, Does Contact Tracing Work? Quasi-Experimental Evidence from an Excel Error in England])
… which would make this Excel error about 500 times “worse” than Therac-25 [Tweet]
(Also, by the way I was roughly this many years minus about two when I learned that arxiv is pronounced archive because of the xi and not knowing this is one of the reasons to dislike professional jargon).
1.2 Griefing
(I am relatively sure I haven’t written about this already, in case I have, I regret the error).
Are you internet old enough to remember the Corrupted Blood Incident [Wikipedia]? Very brief summary: there was a spell in World of Warcraft (the online MMPORPG), back in 2005 when millions of people playing an online game and paying to do so was a novelty and strongly considered to be a portent of things to come. This spell negatively affected your abilities and was also contagious. And lo, there was a virtual pandemic which turned out to be a great event for epidemiologists to study how people might act in a real pandemic.
ANYWAY. I was wondering if the original resulting studies based on Corrupted Blood had any observations on griefing behavior, an example of which would be someone being infected with the Corrupted Blood and then wandering around purposefully infecting people. Because, uh, I guess you can imagine why someone might be thinking about such behavior recently.
(Griefing [Wikipedia] is when… well, it’s basically when you’re a dick to people because you are a dick. In my experience, it’s more like repeatedly killing someone just because you can, and in game environments, rather than the broader behavior of trolling).
Turns out, there was a mention of griefing in relation to Corrupted Blood and COVID-19 pandemic, in a March 13 2020 (OH MY GOD THAT WAS JUST OVER A YEAR AGO) interview between PC Gamer and Dr. Eric Lofgren, an infectious disease epidemiologist [Washington State University]. Lofgren wrote a (the?) paper on Corrupted Blood, The untapped potential of virtual game worlds to shed light on real world epidemics [The Lancet, 2007]
(Aside: I think it’s awesome that we’re at the stage now where people who grew up playing games recognize how they’re relevant to professional work.)
Anyway, Lofgren had this key quote in the PC Gamer interview:
“For me, [Corrupted Blood] was a good illustration of how important it is to understand people's behaviors," he says. "When people react to public health emergencies, how those reactions really shape the course of things. We often view epidemics as these things that sort of happen to people. There's a virus and it's doing things. But really it's a virus that's spreading between people, and how people interact and behave and comply with authority figures, or don't, those are all very important things. And also that these things are very chaotic. You can't really predict 'oh yeah, everyone will quarantine. It'll be fine.' No, they won't." [PC Gamer]
Which, you know, makes a lot of sense to me. Because, of course, not everyone quarantined. It’s worth remembering that this interview was published on MARCH 13, 2020. Lofgren did talk about griefing back then:
“one of the critiques we got from a lot of people, both gamers and scientists, was over this idea of griefing," Dr. Lofgren says. "How griefing isn't really analogous to anything that takes place in the real world. People aren't intentionally getting people sick. And they might not be intentionally getting people sick, but wilfully ignoring your potential to get people sick is pretty close to that. You start to see people like 'oh this isn't a big deal, I'm not going to change my behavior. I'm going to the concert and then going to see my elderly grandma anyway.' Maybe don't do that. That's a big takeaway. Epidemics are a social problem... Minimizing the seriousness of something is sort of real-world griefing." [PC Gamer]’
March 13 was the week things started to lock down across the US and now, a year later, we’ve seen enough videos of people deliberately trying to intimidate others by coughing on them, or flagrantly refusing to take precautions.
Anyway, I think this is a good reminder of two things:
- Games reflect the behavior of actual people
- The behavior of people in games can be reflected in the “real world”
- Games may often amplify the behavior of people given the game’s environment and mechanics.
I’ve written before about Bartle’s high-level player types, which are kind of like a personality assessment that applies to “how people play games”, which point I’ve made before is at least a good provocation to think about “how people behave in “real life””.
Anyway. Griefers. They exist.
And I’m just saying there are established techniques for dealing with griefers in massively multiplayer online games. I WONDER WHAT CAN BE LEARNED FROM THEM AND VICE VERSA.
1.3 Some smaller things that caught my attention’
An expansionist sentient sophisticated monetary exchange?
For absolutely no reason at all, I was reminded recently of Paul Cornell's book, British Summertime, a book about an idyllic socialist future Britain, of green and verdant lands and everybody being terribly nice to each other. Until, of course, a sophisticated, sentient and aggressive monetary system invades and forces them to go to war.
Covers
I love cover versions of songs. There’s something about someone’s interpretation of an original (or preceding) work and then making it their own. There’s the whole Sad Lonely Covers genre, the Epic Trailer genre, and the… plinkly plonky ukele/solo cover genre? Anyway, I discovered SecondHandSongs and it is now one of my Favorite Things.
This is the Second Ever Episode to be published via Buttondown. Things seem to be working? I still need to delete the Substack account - that will be coming in the next couple of days, as soon as I’ve checked that everything is in sync.
It’s nearly “spring break” here in the U.S. which means… it’s the Easter Holidays?
How are you?
Best,
Dan