Ley’s talk from ElixirConf EU is on YouTube now, so go… watch it or something. Here’s the teaser/description from the page:
Phoenix has already proven to be the framework of choice when dealing with a large number of users, on possibly intermittent connections, and all that other well architected Erlang stuff. Which is great when you are behaving as a responsible—-ish adult on the backend of the site.
But now there are literally millions of people coming online, an estimated billion in the near future. These people haven’t seen a web app before, their best connection maybe a household 2g connection, and even their entire country’s connection may have concerns.
I will outline the things we learned, at iwantmyname, about architecting our application specifically for that market and share our lessons learnt while on that journey. This will go all the way from css to how to model your problem in Phoenix to best deliver small consumable chunks.
My feeling is that if the theory is correct that we got into this by trying to be socialist and libertarian at the same time, and getting the worst of both worlds, then we have to choose. You either have to say, “Okay, Facebook is not going to be a business anymore. We said we wanted to create this thing to connect people, but we’re actually making the world worse, so we’re not gonna allow people to advertise on it; we’re not gonna allow anybody to have any influence on your feed but you. This is all about you. We’re gonna turn it into a nonprofit; we’re gonna give it to each country; it’ll be nationalized. We’ll do some final stock things so all the people who contributed to it will be rich beyond their dreams. But then after that it’s done; it’s not a business. We’ll buy back everybody’s stock and it’s done. It’s over. That’s it.”
That’s one option. So it just turns into a socialist enterprise; we let it be nationalized and it’s gone. The other option is to monetize it. And that’s the one that I’m personally more interested in. And what that would look like is, we’d ask those who can afford to — which would be a lot of people in the world, certainly most people in the West — to start paying for it. And then we’d also pay people who provide data into it that’s exceptionally valuable to the network, and it would become a source of economic growth. And we would outlaw advertising on it. There would no longer be third parties paying to influence you.
Because as long as you have advertising, you have this perverse incentive to make it manipulative. You can’t have a behavior-modification machine with advertisers and have anything ethical; it’s not possible. You could get away with it barely with television because television wasn’t as effective at modifying people. But this, there’s no ethical way to have advertising.
The constant mental tug between socialism and libertarianism is something I see in a lot of tech founders. We want freedom through equality, unlimited access, and a public square anyone can participate in. But we also, as Jaron puts it, want to be Steve Jobs.
The answer, as it usually is, is advertising. I’m personally not opposed to an advertising model — maybe I’m in a tech minority. What I am opposed to is advertising based on nearly perfect forms of data.
That probably sounds confusing, so I’m going to come at this as a consumer. I’m perfectly ok with someone like Squarespace advertising on a tech podcast because it’s a fair bargain — I get free content, and Squarespace basically acts as the sponsor. I’m also perfectly ok with reading a food blog and seeing an ad for an awesome kitchen product for the same reason. Enough data exists on a surface level to do the job, and I’m happy that the things I like are being paid for.
What I’m not ok with is looking at a product on Amazon, then seeing that product follow me everywhere I go on the web. And I’m especially not ok with ideas being presented in the marketplace of products — if you’re a 65 y/o woman who goes to church, lives in Mississippi, and posts excited messages about your grandchildren, you shouldn’t be subjected to nationalist political propaganda just because you fit the demographic that is typically swayed by certain types of messages. It’s gross. And it’s bad for society.
So to me, advertising is fine as long as we keep it to products, and we completely eliminate data collection for the purposes of advertising.
So Google has the first model that really works on the internet, which is directed advertising. Where someone basically says, “Hey, I’d like someone to come fix the hole in my roof; I’m in this town, find me a roofer.” And Google doesn’t really want to hold on to you; Google wants to send you on to a roofer very quickly. The roofer wants you to show up as a qualified lead. Everybody’s happy; everybody benefits.
But when I go onto Facebook and I grouse about how much it sucks that it’s still snowing in western Massachusetts and my roof is leaking, I don’t really want someone to lure me away at that point. I particularly don’t want to hear about a vacation; I don’t want to hear about the new car that would make my life happy. What I really want is to grouse and get sympathy from friends, so Facebook at that point is in conflict with me.
And I think what we basically fail to realize is that, in the vast majority of these cases, we’ve sort of picked a revenue model in which we’re in conflict with one another, rather than working coherently in a way where we all end up feeling good about incentives being aligned. For me, the moment where it went wrong was when we failed to realize that there were two kinds of advertising, and that one of them might be healthy for the internet, and the other one probably wasn’t.
Aligned incentives nails it. I want something “free,” a company sees an alignment between their brand and the thing I want, they sponsor the thing. I get my free thing, the people making the thing I want get paid, and now I know about some other thing. Everyone wins.
Nowhere in that process should I have to disclose where I live, what my political views are, who I know, and what websites I’ve been to that day. Stop the surveillance.
Jessica Nordell - How Slack Got Ahead in Diversity
So a team at Slack rebuilt the interview process. For each role, the team determined what characteristics and skills a successful candidate should have—communication skills, say, or capacity for teamwork. Then, for each of these, they defined what information they needed to assess those skills, and then devised a list of behavioral questions expressly aimed at sussing out that information, questions like “Tell me about a stressful situation, and how did it go?” or “Tell me about a change to your code base.” Each candidate now faces the same set of questions. “If you ask every candidate the same question,” Grace told me, “you start to see what a good answer looks like.”
I first read this and went, “Duh, of course you should ask each candidate the same set of questions. How else would you properly compare them?” But then I thought back to every interview I’ve ever run — not once did I go in with a specific set of questions. I consistently make interviews a conversation to determine who — amongst on-paper equals — I would most want to work with. This is a great way to introduce bias, and a terrible way to introduce diversity.
In this age of suspicion — no one trusts Facebook — even asking for what song you heard on Spotify feels like an insult. However, if sharing your musical tastes helps with your likelihood of finding love, chances are you will gladly handover that information. You will even input your physical description, your location, and every goddamn detail about your life. You will leave your intimate thoughts, conversations, and fears on Facebook’s servers, making yourself even more beholden to them.
All this in the name of love! All that data can help them create a better profile of you to sell to their only customer — the advertising community. The brazen audacity of their move is fantastic
“Boss, tracking their every move is making people ask questions. What if, instead of tracking them, we could incentivize them to just hand over all their data. More data than ever before! The possibilities are endless!”
“Yes,” Zuckerberg said in contemplation. “They’ll get married and have babies and tell their children it all happened because of me. Mark Zuckerberg. And then they’ll give us their data, too. Perfect, beautiful data.”
Marco Arment - Overcast 4.2: The privacy update
Overcast has offered anonymous sync accounts since 2014. They’re fully functional, but they lack email addresses or passwords, so they can’t log into the website. A login token is stored in iCloud so the account can be accessed after a restore or upgrade, or from other devices you own.
Previously, the login screen pushed email logins. But with four years of perspective, feedback, and usage data, I now think that’s the wrong move. Only a single-digit percentage of customers use the website, and the iCloud token-sync method solves cross-device logins for almost everyone.
Your personal data isn’t my business — it’s a liability. I want as little as possible. I don’t even log IP addresses anymore.
I’m not a Marco Arment superfan or anything, but this should be up for a Nobel Prize.
I’ll say it again – the best way to grow your business is to make it better. And unless you’ve simply run out of known things to improve, there’s plenty of work to be done that doesn’t require you to be creepy.
Another YouTube description, bleck. The talk is wonderful though.
You’ve read The Innovator’s Dilemma. You’ve bought in to The Lean Startup. You’re ready to kickstart your company’s product innovation efforts. There’s just one problem: you’re not sure where to begin.
What can design teach us about building a collaborative culture? How do you structure a successful product design team? How should that team work? How will that team measure success?
Jeff offers practical, step-by-step, guidance on how to build and support successful product design innovation in your business. Using insights gained from leading and working with dozens of product and design teams, Jeff steps through team makeup, process steps, management structure and the corporate infrastructure needed for these teams to flourish.
Identify a problem, make decisions, ship things, get feedback, internalize feedback, make new decisions/optimize, repeat. That’s how things should work, and It requires, as he says, to build a culture of learning.
When was the last time you got feedback, accepted it, and optimized? When was the last time you listened?
The actual blogroll
(Blogs are ranked in order of appearances in these Blogroll posts. Also, I’m only listing blogs that don’t act as newspapers, because those are the ones I feel need the most support.)
- sparktoro.com (4)
- acolyer.org (2)
- kottke.org (2)
- protonmail.com (2)
- stratechery.com (2)
- om.co (2)
- ncase.me (1)
- thingsma.de (1)
- code.energy (1)
- nomadgate.com (1)
- kapwing.com (1)
- gilest.org (1)
- matthewschuler.co (1)
- someplacestrange.net (1)
- spencerfry.com (1)
- write.as (1)
- zapier.com (1)
- umbrella.cisco.com (1)
- dancohen.org (1)
- pagely.com (1)
- tomcritchlow.com (1)
- austinkleon.com (1)
- smpetrey.com (1)
- resilienturbanism.org (1)
- subpixel.space (1)
- sonniesedge.co.uk (1)
- wormsandviruses.com (1)
- theminimalists.com (1)
- blog.rinesi.com (1)
- cate.blog (1)
- terribleminds.com (1)
- nomasters.io (1)
- randinrepose.com (1)
- blog.evernote.com (1)
- blog.gitprime.com (1)
- marco.org (1)