In 2016 I flew to Dusseldorf to give a talk called Regarding Platforms at B4B DayOne. I just reread the talk track and clicked through the deck, and I want to do a quick lookback because I never posted about it here. Two reasons for this: one, to debut my new homegrown slideshare alternative; two, because I'm on parental leave and as I'm told, you should sleep when the baby sleeps and write when the baby writes.
So. What was I right about? What I was wrong about? What lines would I reuse from 35-year-old Clay?
The argument, in shorthand:
- Bill Gates defined a platform as a thing where the economic value of everyone using it exceeds the value of the company that built it.
- Cities are platforms in that sense – they grow superlinearly because they're made by everybody, not designed by experts.
- Companies, per Geoffrey West's research, scale more like biological systems: maintenance eventually eats growth, and...they die.
- The fix: make companies more city-like by building platforms inside them, lowering the maintenance burden so growth can compound.
What I got right
My favorite Clay Shirky quote is still a keeper. "Communications tools don't get socially interesting until they get technologically boring." I used it to argue management systems were still in their hobbyist phase – too interesting to be useful. Watching AI become boring infrastructure (to some) 2025 and 2026 is exactly this dynamic playing out. The genuinely weird stuff is happening now precisely because the technology stopped being a magic trick.
The diagnosis that maintenance eats growth. Everything I'm doing now at work – the company-as-code database, the org-spine work, the durable workflows architecture – is downstream of this idea. If you can't make maintenance cheap and legible, growth never compounds. Baby-me identified the right problem via Geoffrey West.
I still love that Airbus A380 story. Two centimeters of wiring, two different versions of the same CAD software, two years late, six billion over budget. It's a perfect story about what happens when the substrate underneath a company is poorly maintained.
What I got wrong
As ever! I was (am?) too optimistic about self-organization. The whole talk leaned mutualist – give people autonomy, let networks form, watch emergence happen. A thing I'm working on now called "Two Camps" (coming soon to an inbox near you) says that's half the story. The structuralist tradition would point out that platforms also need governance, standards, and accountability, or they get captured. Which leads to...
Discounting platforms' extractive tendencies. In 2016, my talk track on platforms was pretty utopian. By 2018 or so it was clear that Bill Gates' definition leaves out an important case: platforms that extract more value from users than they return. Uber, Amazon's marketplace, the whole surveillance-capital stack. I didn't reckon with the extractive side, and the talk feels a little naïve as a result.
Platforms not fixing engagement. I trotted out the usual 70% disengaged number and proposed platforms as the fix. Ten years later the number is still the same. Either the prescription was wrong, platforms-as-I-defined-them didn't address the actual cause, or companies didn't really apply them internally. I think I'm still mostly right; engagement is downstream of meaning and agency, but perfect internal platforms don't necessarily touch either of those things (see the point above)...and the primitives described in Jack's Block essay still feel like completely foreign objects to most organizations.
And (of course) It wasn't about computing power, but about attention. The Moore's Law story was correct, but the thing that actually mattered for the next decade was, yes, transistor density and the scaling laws for foundation models, which weren't being talked about much in 2016. I was pointing at the right kind of curve but on the wrong hardware.
Lines I'd still use
- Platforms are only made of people.
- The everything store, supplied by everyone.
- Cities » Ease. Boring » Easy. Platforms » Ease.
- Bill Gates: A platform is when the economic value of everybody that uses it exceeds the value of the company that creates it.
- Jane Jacobs: Cities have the capability of providing something for everybody, only because, and only when, they are created by everybody.
The Jacobs and Gates pairing is the one I'd keep above all the others. They're saying the same thing across sixty years and two industries – value accrues where everyone gets to build, and only then.
The rest of the deck is mostly me being thirty-five and excited about my own micro version of a TED talk. Which, fair.
