s07e14: Dunbar’s Goose and the Pump Handle

Seriously who comes up with these episode titles

0.0 Sitrep

It’s the evening in Portland, Oregon on Wednesday October 16, 2019.

The rain has started here and probably won’t stop for the next few months.

It’s getting dark.

1.0 Some things that caught my attention

1.1 Dunbar’s Goose and the Pump Handle

This is how my brain works:

I was out with friends this last weekend doing some Immersive Theatre (fun!) and afterwards we were having that late Gen-X/early millennial conversation that goes a bit like this:

  • Man, this was fun!

  • Yeah, so fun!

  • Ah, it’s nice to hang out like this.

  • Yeah, this has been really great!

  • We should do this more often.

  • Yeah, let’s!

  • [twelve years, pregnancies, property purchases and major medical incidents pass]

The deal is that these friends live at most about a 15-20 minute walk away and we’ve lived in the same city for nearly 10 years.

It feels like we should be able to hang out often because we live together, but there are lots of reasons why we don’t/can’t. I’m not sympathetic to the idea that is solely because this cohort and later are experts at ghosting each other (although this also feels true), or that they are flakey. I do think there’s something in the idea that late-capitalism/burnout has something to do with getting in the way of getting together. Life is, you know, tiring? And expensive? And if you don’t spend even more time, then it just gets more expensive? And if you don’t have the energy, etc?

Anyway, that should be able to get together because we live close to each other sparked a conversation about how we could treat ourselves as living in a little village of about 150 people, and if you had “mentions the Dunbar number” on your bingo card for this season’s newsletter than congratulations!

I mean, yes, we could treat our immediate geographical location as a little group of about 150 people. Clearly, it’s not that simple otherwise we would be meeting up all the time. For those who aren’t familiar with Robin Dunbar’s number, let me do a quick Wikipedia précis: it’s a 1990s suggestion by Robin Dunbar, a British anthropologist, that humans can sustain or manage about 150 stable social relationships due to our brain’s architecture. The theory is that the size of our neocortext supplies some sort of limit on processing on managing relationships. Some of this theory is backed up with estimated size of neolithic farming villages. Note: this obviously doesn’t mean that groupings larger than 150 are impossible or difficult, just that Dunbar argues that “150 would be the mean group size only for communities with a very high incentive to remain together”.

So! The rough folk rule of thumb is that 150 works because a) our brain’s capacity and b) when you’re under intense survival pressure, it makes sense when possible to band together.

I had a completely different wild-ass theory in that the 150 also works — or is also reinforced — because the potential closeness of the group makes it easy to construct shared rituals and culture. You know, like… a church congregation? But also the idea that there’s nothing like a village water well, where you’d be able to reliably meet all or most of the group, or there’s nothing like the, uh, village bakery, or the Place Where People Do Their Washing and so on.

So far, so good? All this is, is another angle on local community. When it’s easy to get everything drop-shipped and we don’t all have to go somewhere for common or basic needs, then that’s just one less opportunity to bond and reinforce bonds.

I started this with a this is how my brain works disclaimer and if you thought this was all there is to the story then boy am I here to disapppoint or… appoint you.

Because I think huh, communal water well and then I think well yes, Twitter and Facebook are pseudo-communal water wells because they serve as a sort of gathering place and yes thank you for pointing out that this is just like water-cooler conversation if you have the kind of job where there’s a) a water-cooler and b) you won’t get fired if you congregate in groups and take longer than 3 minutes for a break.

I mean obviously Twitter and Facebook are not literal communal water wells because we have a surfeit of opportunities available to us to do the same sort of thing-ish and they’re sort of virtual, abstracted, splintered/sharded water wells with complicated ven diagrams because everyone’s group of 150 (or checks notes and looks at my Facebook friends list: 1,121) is not the same group of 150. Facebook Groups (which appear to be going gangbusters and generally a good thing for those using them, if we pause and put the extremist groups aside) is another place where you can get that nice 150 number.

Anyway. So I think, yes, look, Twitter and Facebook are our new water wells and I should point out that in our in-person conversation I say this somewhat sarcastically, like a sort of ha-ha, Twitter and Facebook are our new water wells, everything is fine as the world burns around us and then…

Apparently saying the phrase “water well” enough times is not entirely like summoning a sort of horror monster, because my brain jumps over to the 1854 Broad Street cholera outbreak, which for me is the kind of thing you learn about in school. For those not familiar, this cholera outbreak is famous (at least, is notorious to me) for John Snow a) mapping the death rates in districts served by two water companies; b) identifying the public water pump on Broad Street at Cambridge Street as a likely source of 616 deaths; c) persuading the St. James parish authorities to remove the pump handle in what would these days be depressingly described as a “This disruptive, life-saving hack probably saved hundreds of people”. This brilliant story was also covered by Steven Johnson’s book, The Ghost Map: The Story of London’s Most Terrifying Epidemic — and How it Changed Science, Cities and the Modern World, which is not only a good read but also has a very long title.

Look, epidemiology applied to online… outbreaks of things is not a new thing. If I recall correctly, people started getting really excited about it when going viral meant some people could get lots of money, which meant people wanted to start studying exactly how they could scientize… (sorry, I meant monetize) going viral as much as possible. I think the first time I heard about the k-factor was in the heyday of social gaming when Zynga and other social darlings of the time like Paul Adams (Grouped, published in 2011), would come by to the ad agency I was working at and say how they had cracked it. And, you know, they had! Look at them now!

But, look. We have ideas spreading and we have extremism and radicalization spreading online and it would be tempting to think that if only there were a pump handle to remove from Twitter and Facebook and maybe the world would be a better place, and I think the answer is: well, yes? That, and probably a whole bunch of other things?

Anyway, we’re only at the pump handle part! I completely forgot while writing this about the Dunbar’s Goose part! This part is actually really easy, and it was more along the idea that you’ve got this super nice village of people but, you know, what brought them together in the end, what really brought them together was this fucking goose doing goose things everywhere. Maybe the real friends were the ones we made dealing with that fucking goose all along.

1.2 No Ambient Computing Without Permanent Surveillance

Ben Thompson over at Stratechery wrote today about Google’s plans for ambient computing, noticing that the world’s self-described organizer of information has clarified its goals and now wants to be helpful which, you know, seems innocuous. Who doesn’t want someone or something to be helpful?

But before we get into this, we should have a brief moment of silence for the now-discontinued little smart-as-a-puppy-camera-that-ultimately-couldn’t Google Clips. A tiny offline, on-device processing A.I. backed camera always on the lookout for “interesting things”? Yeah, I was always interested in that and was just thinking I’d like to try it out. And now it’s dead. May it join the other products in the Google Graveyard and rest in piece.

Back to ambient computing, though. Ambient computing is back! It’s the new everyware, the new smart dust, the new thing that will tie together all of those monetized devices (if you’re Apple) or that will tie together all the data you have about everything, everywhere (if you’re Google).

Here’s the description of ambient computing that Thompson pulled from Google’s keynote:

In the mobile era, smartphones changed the world. It’s super useful to have a powerful computer everywhere you are. But it’s even more useful when computing is anywhere you need it, always available to help. Now you heard me talk about this idea with Baratunde, that helpful computing can be all around you — ambient computing. Your devices work together with services and AI, so help is anywhere you want it, and it’s fluid. The technology just fades into the background when you don’t need it. So the devices aren’t the center of the system, you are. That’s our vision for ambient computing.

Now, I have no idea how Google are going about this internally. I have no idea if there are people like Google’s former Chief Ethicist [sic] raising valuable criticism and asking questions like: what are the failure modes of ambient computing? What does an abusive relationship with ambient computing look like? What are the opportunities for misinformation and deception with ambient computing? Because “helpful computing [that is] all around you” requires some sort of persistent, pervasive surveillance.

The point here is a bit like one I worked on for an op/ed that I never published, but the fundamental point is this: the OKRs for something like “helpful computing” will in some way necessarily depend upon the quality of the predicted services required by the end-user. It is easy to imagine that the easiest way to think you’re going to improve the quality of those predictions is by harvesting more data. And before you know it, we’ve slipped down into another hellhole.

We have a chance, again, to better consider what we-as-society want here. Of course we want helpful computing. But another way of looking at that question is, what are you willing to trade for helpful computing? Mickey got his sorcerer’s power, but are we willing to let other people make the decision about what we’d have to give up?

There’s one example that, unsurprisingly, makes me nervous. Brad Fitzpatrick had a great thread a few weeks ago about Google Assistant/Reserve with Google “causing severe damage to restaurants [and] eroding the trust of restaurant guests.

If you recall, Google Assistant was the thing that would automatically call a restaurant or salon for you and make an appointment because these days we’re terrified of any human contact. There appear to be failures of the system that suggest two outcomes: either Google didn’t think about these things happening (not a great look for a company that prides itself on being super smart) or they did know they might happen and elected not to care (similarly, not a great look ever, and especially not these days).

When a system is supposed to make things easier but hides a bunch of shit under the rug to make it look fine and neither the restaurant nor the user know about “incorrect and incomplete calls” then it’s not so bad for the customers (sure, they lose trust in the restaurant and Google), but it’s ultimately business-ending for restaurants. There’s an interesting point too, which is that most restaurants use the cloud for reservations and don’t use staff to “handle that volume through phone calls.” In Fitzpatrick’s thread, he explains the economics of restaurant phone-answering (it’s not worth it, and you want to do it as little as possible) so just the idea of Google Assistant is apparently one of those things that sounds great on paper and is easily grokkable (nobody likes calling restaurants!) but just… not true on the ground now that OpenTable etc exist.

Again, this feels like a problem a little bit more research could have answered. Easily.

2.0 A short unordered list

Here’s a few things that caught my attention and I promise I really will try to keep it short this time.

Bomber, by Len Deighton, was the first novel written on a computer word processor. When I first saw the tweet by @pulplibrarian about this I knew I had to go and check because you really don’t know if what you read on the internet is true these days. But it is true! And anyway, you should really look at the tweet because it’s got a photo of Deighton with his “91kg leased IBM MT72 [which was] carefully winched into his London home”. Deighton has on the wall behind him maps that he presumably used in the course of writing Bomber, and they even have those bits of string on them. I now feel much better about my Mac LC III.

Nasa has observed an interstellar comet traveling at “an extraordinary speed of 110,000 miles per hour” which sounds like a lot but is only 0.000149c. Space continues to be big.

Because branding, you can (still) get Hummer cologne. I didn’t want to know this and now you know it too. I’m not sorry.

Meanwhile, also via The Drive’s Warzone, Kratos, who make the XQ-58A Valkyrie unmanned combat air vehicle have shown off how it could be transported and launched via an ISO standard shipping container which I bet helps everyone feel more safe and secure.

I had missed my friend Damien Williams’ article last year, What It’s Like To Be A Bot, about consciousness and embodiment.

Epic Games recently launched Chapter 2 of Fortnite by replacing everything with a giant black hole for a few hours and of course a whole bunch of people watched it. I continue to maintain that Epic and Fortnite is much more likely to evolve into the Metaverse because people actually want to spend time there to do things, unlike whatever Facebook and Oculus are trying to do, again. Sometimes it really is the content, stupid.

I’ve been thinking about how I can be more effective and helpful during the day job, especially during meetings. Alan “spoon” Rickman says “you only speak because you wish to respond to something you’ve heard” which I fully support as meaning you should damn well pay attention to people in meetings.


I spent a few minutes today starting to rework the outline of the novel I’ve been working on. Here’s, um, a preview:

HELLO, WORLD takes place right outside your window, ten seconds from now. There’s a pharmacist, tired of helping patients who can’t afford their medication; a high schooler targeted by a radicalized classmate; an assembly line worker in Shenzhen; and a software designer who’s just been told she’s “not a good fit”—again.

Then Silicon Valley’s dream and worst nightmare comes true: overnight, billions of people wake up knowing how to take control of the machines and computers running our planet.

While governments and companies fight to preserve their power, their lives collide as they each take the first steps to remake and reboot the world.

In other news, it turns out the latest version of Microsoft Word won’t automatically open Word files made by Word for Macintosh 5.1a! It’s a blocked file type. So I guess RTF is the way to go.

OK, that’s it for today! Friday’s episode will be s07extra04 for subscribers only, you should be able to click through to the Substack website to subscribe if you feel like it.

I hope you’re well, and as ever, I love getting notes from you.

Best,

Dan