Excited to share that @chdkz is joining @naturalpay today from Notion.
Chika sent me an email on a Sunday night saying that he was on a flight to San Francisco. It was well written, but what I appreciated most was how much of his personal ethos was woven through the message.
Reading it, it was pretty clear that I was going to have a short window to make sure he came to work at Natural.
Meeting him emphasized his competence, but what stood out was his humility, his commitment to being excellent, and his willingness to bet on himself, all without ego.
There are very few people who come through the office so capable, with such genuine curiosity, and real sincerity about the process.
There is no question that Chika moves through the world incredibly intentionally, and with an insanely steep slope. If you interact with him, you’re struck by just how capable he is in such a short amount of time.
By the end of his work trial, there was no question about whether we were going to try to hire him.
Excited for him to officially make the move from New York to join the team.
https://t.co/RLzHdlmi7j
“But the solution when a set of facts conflicts with your existing beliefs is to think harder about them, not operate in realities in which those facts do not exist.”
Loosely reminds me of: https://t.co/AhKX6ZO7Ig
As I've gotten older, It's been quite fascinating to watch how some of the opinions/takes I once held so strongly are now significantly updated or totally discarded altogether.
And it's usually not the case that the original take was flat-out wrong, it usually just tends to be *incomplete*, missing a fair bit of nuance -- typically because I didn't have enough datapoints relative to ${now}.
This has been consistent enough that I can extrapolate to know that future me will prob look back at current me in a similar way for some things.
So, I'm learning to factor in my blindspots, i.e. the "unknown unknowns", whenever I'm forming an opinion or making a decision; as I know that I'm almost certainly missing some critical datapoint(s), which only time + more life experience can uncover.
> Was Cursor through and through because of the intuitiveness behind a GUI
I was recently thinking about this and, for me, idk if it's more about intuitive GUI as much as it was about "trust." Earlier on, say pre Opus 4.5, I pretty much used Cursor exclusively. This was because I didn't feel super comfortable using CC to make a bunch of edits across a bunch of files in one go. Cursor's interface was much more appealing to me cause it felt easier to review stuff and accept/decline or modify changes granularly. Basically, in a world where the models were "pretty good, but still required a bit of handholding", I felt that Cursor provided an infinitely better experience than using say CC + IDE.
This became kind of a non-issue In the post Opus 4.5 world, where I started to feel like the models were much more reliable in one-shotting stuff. At that point, the review safety net mattered less, and CC's speed/simplicity seems to have won out.
Tbf, looking back at the JSON file in my text editor, I guess you _can_ tell them apart and it's not as similar as the ones in Claude's response
But like, realistically though, this is an extremely easy thing to miss when debugging in the heat of the moment.
I just spent hoursss trying to figure out why an Athena query was returning 0 results, even though the data was clearly in the relevant S3 bucket.
Turns out it was cause of one char -- a rogue curly quote hiding inside a JSON key.
Man, I thought I was going crazy...
The loudest story about AI is a lonely one. One person with an army of chatbots. Other humans are friction.
That gets the future wrong. The best things aren’t built alone.
In a moment of change, we want to remind the world (and ourselves) what Notion stands for:
— Think Together