Hacker News new | ask | show | jobs
by jgon 4721 days ago
Reading this article the one idea that persisted in my head was that this was a case of people not really knowing anything about ecology trying to create an ecosystem. Sure they used the word "Platform" and maybe that was actually the first clue, but what they really wanted was an ecosystem of actors creating increased value for their domain.

I think that my favorite piece of writing by Cory Doctorow is his essay "All Complex Ecosystems Have Parasites" and I thought of it early into this piece. Any ecosystem of any interest will always have parasites within it. When you attempt to indiscriminately remove them all you almost invariably end up with something sterile and far less interesting. When I discover a weed on my lawn, my reaction isn't to burn the lawn and pave it. Sure I wouldn't have any weeds that way, but a concrete sheet is far less interesting (to me at least) than a lush lawn. So I suck it up and deal with the weeds on a case by case basis, because that is the only way to do it without torching the rest of the ecosystem that I want to maintain. It seems like Facebook, when faced with the weed problem, went the concrete sheet route, torching both the parasites and the rest of the actors that actually made up the ecosystem. Sure, it means that there are no more weeds, but it means that there also aren't any of the other organisms that create value in your ecosystem.

Another thing to consider is the energy cost of "putting down roots" so to speak in an ecosystem and evolving with changes. An organism has to spend energy to establish itself within an ecosystem, and it also has to spend energy changing itself to adapt to changes in the environment. Every joule of energy it spends doing this is a joule it doesn't spend enriching the ecosystem it is a part of. Going back to my lawn, I don't change the soil composition every year, and I also don't pull up the grass and set it back down to resod every 6 weeks. Eventually it would die, because I would be forcing it to spend all of its energy adapting to changes I introduce rather than helping the ecosystem thrive. And then I'm back to my barren field problem.

I don't have a good ecosystem analogy for competing against their developers, but I will note that it can be framed in energy terms again. Any other organism, after watchers other repeatedly expend their energy in a particular ecosystem only to be crushed will likely decide against attempting to establish themselves in your ecosystem and instead choose one of several other competing ecosystems, even those that appear initially less attractive, because what is the use if you likely end up failing anyway?

So, my takeaway given only my shallow knowledge of what happened based on this article is that other organizations attempting to establish an ecosystem (and that is really what they want when they say platform) and reap the numerous benefits that an ecosystem can produce should spend some time thinking about how ecosystems become established and thrive in the real world before charging clumsily forward with their own naive attempts.

2 comments

This is an overly simplistic analogy.

No platform developer, when faced with weeds would say "let's torch it".

What really happened in my opinion is that they tried a weed killer. However, as in real gardening, it turned out that distinguishing a weed from a non-weed is complex, subtle, and full of value judgments. In some cases, even if a plant has weed-like traits or capabilities, it's not a weed in the eyes of the gardener. It depends how the capabilities are used.

There were also cascading effects (and here is where I really stretch the metaphor). Once a few non-weed members of the ecosystem were killed, the garden suddenly looked a lot less inviting to others.

I agree that the analogy is simplistic but I disagree with your weed killer analogy. I use weed killer on my own lawn and when I do I apply it in a targeted spray specifically to the weeds in question. I am still making a judgement and doing the hard work to separate parasites from useful organisms. Based on the article is sound like they sprayed everything ie removed the apis in question, and weed killer will indeed kill the grass surrounding the weed if you are not careful.

So I think that the analogy still stands, in that if you want an ecosystem there isn't really any way to mechanize it yet and you have to be prepared to do some of the work by hand. Creating an ecosystem worth having means offering up interesting capabilities. Offering up interesting capabilities means that parasites will appear. You can remove the parasites either by removing the interesting capabilities, or by doing the hard work to single them out. Doing the former can have grave consequences for your ecosystem even if it appears easier/cheaper in the short term.

Simplistic or not, I think the analogy serves its purpose very well (thank you for sharing it). As I was reading the GP post the analogy I was thinking of was spraying weed-killer indiscriminately everywhere and naively thinking it would only take out the 'bad-guys'.

Doing it properly does involve understanding both what 'weed' means to you and 'desirable plant' then being careful not to poison the latter (and actually encourage it) while you try to deal with the former.

I interpreted it as: they removed an API (or reduce the functionality of an API in a way that) they thought was only useful to weeds. But it turned out that there were legitimate uses for that functionality, and it hurt more than just weeds.

I don't really know the details; this is just reflecting my experience from working on platforms.

As you say, in many cases, it ultimately requires human judgement, which is something that engineers are loathe to rely upon.

The cost of parasites was too high for facebook, when weighted against the opportunities: facebook is all about being flat and appealing to the general public, complexities like possible scams are putting off too many potential users. Facebook could not have grown to its present size if it had not killed the platform.

Also they can't share the social graph; the monopoly on the data that they have gathered on us is the whole point/worth/wealth of the company.

If facebook or google run into financial problems (this might happen if tax loopholes are closed, and countries decide that even big companies have to pay real corporate taxes) then these companies will be bailed out as they are 'too big to fail'.