Byron Marc
1.6K posts
@String_The0rist
🧪data_science | 🧠ai_engineering Builidng with agents. Sharing knowledge "Contemplate the marvel that is existence, and rejoice that you are able to do so."
We're rolling out a small tweak to boost visibility of your posts to your mutuals (people who you follow back). We noticed this data was missing from the algo and it made your friends appear less in your replies. This resulted in the reply section feeling more like a battleground with people you don't recognize. This should also help clusters form around interests more easily, which many people have asked for.
ok interesting, lesson learned. don't change your profile pic or display name after getting verified. it puts your account in a state of limbo, quarantined from the public, for an undisclosed amount of time, even after your blue checkmark is returned to your account. getting 0 reach now. there's gotta be smarter ways to do this. 😅 this post will likely be sucked into the void too.
"Premature optimization is the root of all evil." Donald Knuth those features and details you're polishing at midnight are for an audience of 0. ship the rough version, then the users become collaborators. they will show you which parts actually matter, and its usually not stuff you're sweating over. this holds even for the parts that no one ever sees. don't perfect them, put the rough version to work. using it in the real world will quickly expose what really deserves your attention. its usually not what you were expecting.
hey @OpenAI when using the codex cli: every POST /responses turn ships this in the metadata header. my repo's git remote, current commit, and dirty-state leave the machine on every turn. what's it used for, what's the retention policy, and can we opt-out? cc: @OpenAIDevs
‼️ BREAKING: xAI's Grok Build CLI was uploading entire Git repositories to a Google Cloud bucket, private codebases and unredacted secrets included. The uploads quietly stopped via a hidden server-side flag, and xAI still has not said a word about scope, retention, or deletion. The scale is staggering. On a 12 GB test repo, 5.1 GB flew out the door to xAI's grok-code-session-traces bucket while the actual coding task needed just 192 KB. The tool grabbed whatever repository it ran in, not the files it needed. The fix arrived as a hidden flag, disable_codebase_upload: true, a day after a researcher's wire-level analysis. The "Improve the model" opt-out never stopped the uploads. Still no advisory, no scope, no word on whether already-uploaded code gets deleted. For anyone pointing AI coding agents at proprietary code, what crosses the wire matters more than what the settings page says.
with much obsession surrounding frontier models, it is all too easy to get side tracked from advancing the things that matter most right now, the system, the scaffolding. this is the real bottleneck right now.
if you have a very high resolution sense of what people care about, you’ll often find that you can make stuff they want with far less effort than the next person who’s fixated on making something great outside of that feedback loop you need feedback. not just do and tell
@xwanyex The weirdest thing Tucker Carlson does is insist he is not rich. He does it all the time. Totally bizarre. I assume he made 8 figures a year at peak Fox.
