Meta Meta Meta

[raw note: spent ~30m writing it]

In video games there is this concept of a “meta” - i.e., the “way to play the game best”

If the rules of the game change, the meta changes — meaning the agreed upon standard moves you take to win change.

This happens outside of games too. There’s this tendency for things to start looking the same

  • youtube - mrbeast, other top youtubers have the same style of thumbnails, types of videos
  • social media - meta, linkedin, twitter, they all have the same features and growth tactics

and list goes on for cars, musicals, how to hit a hockey puck, how to box

I wonder why.

Is society naturally optimizing to a global max like some kind of collective organism?

Like, say a new piece of tech comes out and we have humanoid robots

Are the businesses that capture value on that piece of tech structurally destined to look the same because now there exists some globally-optimal best way to make a humanoid robotics company

I think so. It makes sense that there is “some best way to do things” or perhaps not only 1 way but a set of ways.

And I do assume we roughly gradient-ascent to this optimum, and that incentives are this underlying thing that govern us — effectively social laws of physics.

But I think despite some few people innovating, most are copying the status quo

So interestingly if 10 robotics companies launch after a new piece of technology release, I assume 8 of them are primarily just copying the other two how they grew, and 2 are really trying to innovate on something

Paradoxically if people do behave this way and just copy, we won’t find a better optimum.

So how do I balance this?

Imagine I’m making my own new Meta — like the company that made facebook

So they established what I’d call the meta-meta, i.e., the “typical way one makes a social network”

I want to know the strategy for in such cases to what degree I use the meta, or define my own

I want to come up with a Meta Meta Meta

I know this could apply to other things than a social network, but I like saying Meta 3 times so I’m using facebook as the example.

My meta meta meta

  • if I were starting my own facebook i’d first copy exactly what they do to hopefully get ahead quick with minimum thinking - so I can just focus on moving fast and executing.
  • As I do this I’d dissect how and why facebook is the way it is, try to reverse engineer it down to the core principles that govern the design
  • i think from doing this I’d realize some decisions were just arbitrary - some were wrong - some perhaps were correct for the facebook team but not given my team or starting conditions
  • so now I’m in a better position to figure out what things I want to do differently vs copy
  • now there is a lot of work to experiment - I’d just apply the scientific method to come up with bets on what will lead to a better social network - try those things out - learn really fast
  • if I do find a way to make a better social network - now I need to capitalize on the advantage
    • e.g., in a video game I’d win a lot. in a sport I may win olympics like dick fosbury with his first-time being someone to back-flip over the high jump bar, in a company I may dominate a market for a while
  • now I’d need to lock in a structural advantage to stay ahead

    • This matters mostly in capitalism I think.

    In the olympics for example if I come up with a new technique and win then I’ve already had a great career, i don’t think others can copy me in a way that holds me back - if anything I get to define a new meta for a sport and it elevates my status as an athlete even more

    In building a company though I’d need a way to stay ahead and keep making money. Like network effects.

    One option could be to just keep innovating faster. Have the meta keep evolving.

    I like this one as it seems the coolest, and it may be true as technology advances faster and faster structural advantages matter less as you can just be on the front; but i’m but unsure how easy it is to do in practice. Also, staying ahead because you’re already ahead is a form of structural advantage - i.e., first mover advantage - people want to join the winning horse - which may help the horse keep winning.

    I think this is what OpenAI is doing essentially in 2024 - they don’t have much of a moat, but they are winning - and so smart people want to join them - so they keep winning.

    In the case of AGI if there’s no cap then first mover ends up being last mover and effectively conquers the universe.

    If there is a cap though, now I’d hope I had some structural advantage so I can maintain my position at the top.

To follow the meta meta meta, you must first be fairly immune to Prescriptive Anxiety or else this may lead to you zapping creativity.

I welcome comments + discussion @ kylemorriscs@gmail.com - I prefer 1:1 so discourse is less mimetically influenced