The blank slate in your hands
How do we rebuild the internet from scratch?
A couple of housekeeping things: I’ve been pretty busy lately!
I gave a keynote at RStudio’s 2021 Conference. You can find all the slides here and it’ll be up on RStudio shortly!
I gave an interview for Coralie’s Parenting in Tech project project
On to the newsletter!
Portrait of a Young Man, Antonello da Messina, 1478
What would you do if you had a second chance to recreate the internet? That’s the question I find myself asking all the time lately.
We now know about Cambridge Analytica, about the Pinterest lawsuit, about Project Dragonfly, about Google and the Nothing, the $1 billion Rubik’s Cube, the Experian Data Breach, the Google AI controversy, the Whatsapp thing, the other Whatsapp thing, It’s Time to Build, and the 10k people in San Francisco. We know that humans are not Web Scale.
We messed up.
Where do we go from here?
Let’s say you have a ground floor shot at being at Facebook. What would you have done differently? For me, It’s obvious that I wouldn’t make it about advertising, and I wouldn’t do mass data collection or try to mandate real names.
But if there was no advertising, there would be no money for Facebook, which means it would never survive as a company. Which means all the good it’s done like connect people with rare diseases to share information, form parenting support groups, and connect small businesses with the community would also be gone, too. It wouldn’t have connected me with some of my best friends. Facebook is more than a misinformation machine, it’s also a marketplace, a place to share pictures. I can tell how vaccination is going in my area by checking Facebook.
We live, as always, in the age of the great cross-hatch. Social networks are largely this big, horrible thing, but they also have very real positives. How can you tell, when you’re starting a company, which of these things the company will be? Will your humble notetaking company, Evernote, be a place to store all of humanity’s knowledge, or will it be an awful arbiter of secrets great and small in case it accidentally opens everyone’s notes. Will your work communication platform, Slack, be a positive boost of productivity, allowing people to work remotely, or will it be this huge octopus, tentacles reaching through all your devices to make you work off the clock?
We don’t know until we file the articles of incorporation and set forth.
Sure, we can guess. We can set up guardrails. That doesn’t at all excuse thinking these things through as we work. “It’s hopeless, so we shouldn’t even try.” But, there is a compound tragedy in this wonderful, terrible last fifteen years in technology. The tragedy is this: we humans are terrible at predicting things. (Don’t tell the data scientists.)
We can asymptotically approach reality, to an extent. But, all models are terrible, George Box said. We don’t even fully understand numbers. And we don’t know for sure what the future will hold. And, what’s worse we are terrible at understanding how systems will interact. If we decide to track users, what will happen and what won’t? If we decide to track some users? We are, in a way, always trying to understand the beginning and end of this meme:
How do you get from Zuck trying to start a social network to pick up co-eds to the fall of democracy? And what happens five years from now? We don’t have big enough of a neural net to model the predicted possibilities.
And the second tragedy is that all companies must fight the market to survive. Companies start as babies, when the odds are entirely stacked against them, and, at the beginning, they don’t have the luxury of making decisions based on ethics or guidelines or frameworks. They are hungry and blind, and they must go after the money.
If you don’t believe me, read Shoe Dog or The Making of Prince of Persia or The Great Beanie Baby Bubble or Skunkworks or Super Pumped or any one of the foundational books about how companies grow and thrive. There are some that take the very hard path of pursuing ethics first, but at the cost of growth later. Almost no one, especially anyone working on money borrowed against them, can afford to do this.
This is not an excuse for big companies acting like marauders. But it’s just an observation that most companies don’t start this way that we want them to, with a balanced deliberation. They start in a desperate panic to survive, with a bunch of compromises that they forget, with un-archetected data, with “we’ll fix this later.” because if they laid all of this out at the beginning, they’d never get off the ground.
What do we do about this, these two fundamental laws of physics and capital? Knowing what we know now, how can we fix the internet going forward?
I’m not a prophet, just a rando with a newsletter.
But it’s 2021, and we’re sitting at our desk.
The VC calls.
“It’s time to build,” he says.
What will we say?
What I’m reading lately:
Interview with Max Levchin. (Previously in Normcore on him)
Artificial intelligence is a house divided (AI previously in Normcore)
Congress basically says YouTube needs to shut down its recsys (YouTube recsys previously in Normcore)
The Overstory (it’s good! There are trees!)
The Newsletter:
This newsletter’s M.O. is takes on tech news that are rooted in humanism, nuance, context, rationality, and a little fun. It goes out onceish a week. If you like it, forward it to friends!
The Author:
I’m a machine learning engineer. Most of my free time is spent wrangling a kindergartner and a toddler, reading, and writing bad tweets. Find out more here or follow me on Twitter.