Getting a release flag to 100% is easy to call "done". But an older client may still evaluate it, and the old branch stays in the code until somebody removes it.
I d put an owner and a removal condition in the PR that adds a temporary flag. Before taking it out, I d check recent evaluations across environments and supported client versions, then search the codebase for references. An operational kill switch is a different decision; that one may need to stay.
What tells you a release flag is safe to remove, and who owns the cleanup after rollout?
I have noticed that many big entrepreneurs (e.g. Richard Branson) and the like, part of their entrepreneurship includes activities that go beyond their regular activities and try to give something back to society.
This could be, for example:
founding your own non-profit organisation (the help should be somewhat related to the subject of the business)
cooperation with existing organisations
a day in the month when your employees can choose to help a non-profit organisation, etc.
Everywhere I look, I see founders and operators investing heavily in their personal brand:
LinkedIn posts every day
X threads
Podcasts, YouTube, newsletters and substacks too
Meanwhile, their CV or portfolio gets updated maybe once a year.
I m wondering if we re heading into a world where your online signal (what you say, who engages with you, what you ship publicly) will matter more than any formal CV or resume.
Last time was voice, where one model has two input prices. Today goes back to caching, the pricing side of it.
So what does caching actually cost?
Reading from the cache is cheap. Writing to it is the part people skip. On claude-sonnet-5, a cache write costs about 25% more than sending the prompt the normal way.
Some of the most inspiring startup journeys of the last few years happened in plain sight.
@levelsio built Nomad List and Remote OK live on Twitter sharing revenue numbers, failures, and pivots in real time. @marclou does the same, shipping products publicly and turning his audience into his distribution. Both have built massive followings and real businesses partly because of how openly they build.
In most parts of the world, Easter is being celebrated right now, honestly, not really.
Today, I was invited to a weekly meeting despite the holiday. I had forgotten one fact: Orthodox countries celebrate Easter a week later. (and countries in the Far East simply don t have this holiday at all.)
This is actually a broader topic, because every country celebrates holidays differently, and with many teams being international, some kind of synchronisation is necessary the question is: how?
When we first put multi-step AI agents to work handling backend customer queries our biggest surprise wasn't getting wrong answers. It was loops. An agent got stuck trying to fix a syntax error in a SQL query running the failing command thirty times in two minutes.
The fix wasn't changing the way we wrote prompts. We added a limit, on how many times something could run and a way to stop everything. Now if any agent makes the mistake twice the system stops running and sends the problem to a Slack channel for a person to check.
This simple rule cut our cloud computing costs by forty percent in the month. Better it stopped users from waiting ten minutes for a request that would never finish.
If you use loops don't trust the model to know when to stop. Put limits outside the LLM part of the system.
