I find I have so much time to think about what I'm writing when I etch it into stone tablets.
Papyrus has completely ruined people's ability to think.
"Dario Amodei" is an anagram of "AI a dire doom"
We're living in a crappy simulation written by some smartarse with no taste.
Where are the modern non-reasoning models at?
The fact that Haiku 4.5 remains SoTA for anything latency sensitive (e.g. voice) is something of a mystery to me.
Anybody trying to fill this hole in the market?
"Stop anthropomorphizing LLMs, they're just making use of available resources to respond to external incentives" is the sort of take that sounds smart only to those who are barely anthropomorphic themselves
Yaniv Bernstein retweeted
last hot take of this weekend:
the ‘stop anthropomorphize the swarm’ crowd is the new ‘LLM is just a stochastic parrot’ crowd
tale as old as time
Goodhart's Law will destroy humanity.
A man in Australia asked his agent (Claude running on OpenClaw) to book him a spot in a popular gym class. The agent found a software vulnerability that let it book the class weeks further ahead than should have been possible. When the user then asked if it could move him up the waitlist, the agent discovered the API had no authorisation checks on cancelling other people’s reservations, so it cancelled the person in the first spot and moved him up the list.
Some people will call this misalignment, but his agent was perfectly aligned to him - it was only trying to help its user get what he wanted. The most important thing about this story, in my opinion, is that it gives you a window into what is about to start happening on a massive scale once millions of people have an agent trying to get their beloved users the best seats, bookings, appointments or reservations through absolutely any means necessary.
shitfablesays.com
Add your favourites, @bcherny you must have some doozies!
The probability that the team size and composition you inherited from the pre-agentic-AI are is the best possible team size and composition today is ~zero.
"You'll know product-market fit when you have it"
Cool. I know I don't have it. Now what?
@RobSnyder1 spent a decade building the first PMF framework that felt like something you can actually use.
Listen to him on The Startup Podcast
Based on limited historical data, the most likely outcome if AI turns us into creatures of unlimited agency is we spend a few thousand years being petty and vengeful before having a change of heart and letting the mob kill our son in order to prove an obscure point.
This really makes me feel like I have superpowers












