Episode Eighty Four: Day One; Google Everything

by danhon

0.0 Station Ident

I write this at n thousand feet at x speed, on my way down to San Francisco for interesting meetings and O’Reilly Solid, of which there seems to be enough palpable excitement around for it to feel just a little bit like the O’Reilly ETech when it was good, which was the first one, which I wasn’t at, not the second or third ones which I can’t remember which one I went to. But anyway. Lots of interesting people in San Francisco at the moment, and I’m looking forward to meeting them and getting a little bit (or a lot) of friend time and gentle ego massaging.

1.0 Day One

Things I haven’t done:

– broken down and cried (though I’ve felt like it)

– broken down and ordered delivery pizza (though I’ve felt like it)

– replied to all the incredibly kind emails I’ve received

– had a good night’s sleep

– had too much pride to ask people for LinkedIn recommendations

– not participated in drunken karaoke

– given up

It already feels different. I dropped off the severance agreement today, signed, sealed and delivered. I have some interesting meetings coming up, for which I thank you all, but it never hurts to have more, so if you’ve been wavering because you think it’s going to be super-easy for me to pick myself up and get out there, then don’t waver: send me those introductions or get in touch. It all helps.

One thing that hit me this morning as I was busy doing things like filing for unemployment, finishing a load of laundry, packing up for this trip down to SF, feeding the cat, picking up a new trike for my son from Craigslist, was: my self feels remarkably functional for what feels like has been thrown at it. (It probably isn’t a good sign, though, that I’m able to disassociate from my self though, right?)

I know – I mean I *know* know – that only a few months ago, if I’d been in this situation, alone, with the past few weeks that I’ve had, I’d be a pretty dysfunctional wreck. And yes, I have new drugs. And bonus drugs on top of those. So thanks, pharmaceutical industrial complex, for tweaking my neurochemistry in the required direction.

But I think one of the other things is that I’ve been writing this every day.

I knew – and remember talking to my therapist about it – that the time would come when my mother-in-law would pass away. And I was worried that I wouldn’t be able to or couldn’t write, and that what’s happened with that before has been the feeling of all-or-nothing, of having gone up eighty-odd steps, every single day, and each day accomplishing more and then having to lose it all with one mis-step. Because the way my brain works, one failure is – or maybe was – good as total failure. And that one mis-step would be so demotivating, so crippling, that it would turn into another mis-step the next day. And the day after.

But when I sat down to do it, every week day, all I have to do is just write. I just open up a text editor and start typing. And I don’t care if it’s good and I don’t care that it’s bad and I don’t even necessarily care that you’re on the other end of the SMTP connection.

But whatever the writing’s doing, I think it’s doing a good thing. So I’m going to keep doing it.

2.0 Google Everything

If you’ve been on the internet for a while, then it’s likely that you’ll have spent at least some time on Metafilter. One of the prototypical community sites, Metafilter is of a web 1.0 era: I remember when it updated to include fancy AJAXy features like having new comments append to the page instead of you having to reload the entire thing.

I’m lucky enough to know Matt Haughey, Metafilter’s founder, and to have spent a fair amount of time with him since I’ve moved to Portland. Generally, we go on geek dates like when he DM’d me and asked if I’d be up for going to the auto show and laughing at bad car user interfaces.

Anyway. The thing about Metafilter is that it’s a community site. And it’s a community site done the hard way, in the way you used to read about in books like Powazek’s Design for Community[1]. The hard way, of course, is by almost excessive use of humans. And not just any old humans, but freakishly empathetic humans: ones who are able to magically discern just through TCP/IP packets and their understanding of human behaviour, what it is that people on the other side of a screen and keyboard actually mean, as opposed to what they typed.

Google’s AdSense and its counterpart, the search algorithms and pagerank that power their search service for inbound and organic traffic, proved (notice the use of past-tense) an excellent way to fund the more idiosyncratic (and less, I supppose) parts of the web. Especially the useful parts. But now it looks like that’s changing, and in the same way that the newspaper industry has its own church and state divide, Google has its own with the search product team and the AdSense product team.

The crux of the matter is this: that Google ad revenue peaked, and – in the conversations that I’ve had with Matt – part of the issue is that the church/state wall is that it isn’t a wall – it’s a permeable business membrane, one that is algorithmically defined in a way that, I suspect, cannot be predicted or modelled in terms of its real life impact. And at the heart of the crux is something even simpler: what type of content does Google’s pagerank algorithm favour, because the truth is that we have a content ecosystem that *in some pockets* (cue: Not All Googles) has a bona-fide monopsonist.

You could look at this another way and say: well, Metafilter got fat on the teat of Google and should have made sure that its revenue came from diverse sources. Whatever. We’re here now, and the question for Matt and Metafilter is: what now, and what other sources?

To say that Metafilter was a great experiment is to miss the point, I feel. I don’t think it was a great experiment: it was calculatedly tended. You get parks and you get gardens and then you get obsessively tended gardens that are intended to be something of a shelter in amongst all of the noise. In the continuum of community – we can drop the online prefix now, I feel – we’ve had Metafilter and sites like it that acted as our backstop, our hardly-any-algorithmic aids and pure, human brute force. And at the other end, we have our Reddits and Hacker Newses which show us what’s possible when algorithms take a little more control.

This isn’t just curation. Curation’s, I don’t know, level one of community – just picking the thing that your community has surrounded itself with and making sure it stays on track. No, what Metafilter did – does – with its moderators was *direction*.

I feel like I speak with a little authority on this because of the time I spent, a long time ago now, helping to moderate the Cloudmakers mailing list for the A.I. alternate reality game, The Beast. Between the seven of us, we read and, if necessary, responded to *every single message*. And we were strong, and firm and when something happened that needed a light touch we gave one and if it needed a harder response, we gave that too. Because there’s something to be said for leadership, and that’s what Metafilter has: strong human leadership. And good humans cost money.

Without a doubt there’s a lot we can achieve with humans volunteering their time. But that type of involvement requires as much leadership and as much direction because the things we make are opinionated: they reflect the choices we make, and Metafilter in its usage of continual human contact *is a choice*.

If we’re interested in a diverse internet, then we should be thinking about what the evolution of the Metafilter is that preserves what makes good human involvement good, as well as making the best use of automation.

So it’s rather unsettling that news has leaked that Google’s looking at providing a unified CMS publishing and advertising serving platform. With this component, you don’t really need anything else: it’s a one-click company town where you can just tap a card and start your publishing empire. It is, almost, a kind of pro-Blogger, the true democratisation of publishing, but this time a porous church/state divide that allows for content monetisation from the get-go.

At this point, you start wondering (and if you haven’t been wondering, know this: you should *always* be wondering) – how does this fit with the rest of the ecosystem out there? This isn’t just a knee-jerk anti-valley reaction, it’s a genuine question as to how all of this stuff fits together. Is a certain type of content being optimised for? In other words, as people like Anil Dash have pointed out, this is what happens when you try to build heuristics that aren’t flexible enough and you get second-order effects. Sure, linkfarms are a bad, but they make a lot of money for people, and in the race, it’s easier to make linkfarms than it is to detect them. And in wars there will always be collateral damage. In this war, it looks like Metafilter might be wounded, and we’ll know in the next year or so whether mortally so.

[1] Design for Community – Amazon: http://amzn.to/1lD959B

[2] http://www.adexchanger.com/publishers/google-explores-a-unified-cms-and-publisher-ad-platform/

3.0 Requests

I’m going to try an experiment this week: I’m going to take requests. I’m going to have a lot to write about, I expect, what with Solid going on, so this is just my masochistic self peeking through, but let’s have a go at this. Send a suggestion for something you’d like me to have a think and a write about. Some of you have done that already; I’ll add those into the hat. I’ll take a look and pick some to write about on Wednesday, Thursday and Friday this week.

That’s it for today. See you on the other side.

Best,

Dan