in B2C, friction during onboarding is often necessary in order to deliver value
a completed onboarding to only churn right after is worth nothing
this is often jarring when demo’ing your app, but you’re not optimizing for demos, you’re optimizing for real users
Ollie was one of the most interesting vertical AI agents out there for a while.
From their data, the number of external apps a user connects directly correlates with W4 retention:
0 -> 3.5%
1 -> 18%
2 -> 40%
3 -> 52%
That's exactly why we're trying to build for believers first at Quest. People who are willing to connect multiple sources.
We'll probably lose a lot of users initially because of that friction. But that's fine.
It will allow us to listen to the right users, keep iterating on making the product better (because, yes, we know we have a lot to do), and ultimately I'm confident that more and more people will see the value in opening up more to their agents.
Meanwhile, we'll stay heads down building for the right users.
the last 10-20% of a feature still takes 80-90% of the time
building is definitely not a bottleneck for us anymore, but crafting a good experience is not a solved problem with AI
Christian Kjær retweeted
Americans are tired of being told their needs are unrealistic by politicians who have been bought by the wealthiest people in the country.
i just don’t understand why people keep thinking Penix is a viable QB for the Falcons
he’s too stationary making him predictable, too injury-prone making him take less chances, and just hasn’t proven it at all yet
gross misuse of power by Meta locking out other agents, the EU will probably be the only ones that will crack down on this
monorepos are 100% without a doubt a benefit for agents
type checking across your entire stack, the knowledge of how everything interacts, and atomic changes across the entire stack and infra
When I first told @MParakhin that we were planning to move to a monorepo he said wisely:
Moving to a Monorepo is an idea that every engineer absolutely hates, and that once its there, no engineer would ever want to work at a company that doesn't use one. We will definitely do it.
REST is for spaghetti code, GraphQL is for structure
enforcing structure on agents is good
therefore GraphQL is good
got my boy Claude to go ham on the rust graphql library we were using (async-graphql) because the compile times were insane
not only did it reduce compile times of cold builds by ~40% and rebuilds by ~80% but also reduced allocations by ~60%
zero runtime regressions 👌
shared the learnings with upstream, most of the gains didn’t require any breaking changes at all
github.com/async-graphql/asy…
crazy if falcons are the ones to stop the 49ers win streak
this team is really all over the place lol
Airbnb was (and often still is) getting access to someone’s private home to stay in
Leaving trash in someone’s home is weird
But many hotels etc are now on the platform, so the service expectations have shifted, and now the experience varies extremely from place to place
what a weird uninformed opinion to state publicly?…
What’s happening in Paris and Milan is no surprise to anyone who has spent time around young people in Europe during the 2010s:
Virtually all of them were working on their 4th masters degree in social welfare, were unmarried, attended communist dinners every night, and regularly participated in riots.
Now the economy is suddenly transforming and they have nothing to show for it.
wishing someone put in as much effort into speeding up the Rust compiler as we are getting in typescript right now 😭
if buns aot delivers then I go back to full ts for everything (right now we mix rust and ts)
Could not have come at a better time 👀
Replying to @deepseek_ai
🔍 Search & Fetch APIs are now open to everyone!
Powered by our in-house search engine, serving hundreds of millions of requests daily worldwide.
💰 Each endpoint: $0.75 per 1,000 calls.
📖 Docs: api-docs.deepseek.com/guides…
📌 Search & Fetch calls via server-side tools in DeepSeek Harness, Claude Code, and Codex are currently free for DeepSeek API users. Billing starts Oct 31, 2026, at 16:00 UTC.
3/4











