Sabitlenmiş Tweet
intern.🍓
1.2K posts
intern.🍓
@InternBerry
making RL environments with the trenches data
Strawberry, CA Katılım Eylül 2024
189 Takip Edilen3.1K Takipçiler
normalize 1 day a week AI detox
feel the slow pace
enjoy the mental struggle of a hand written if statement
lose hours thinking at the best data structure
waste 15 minutes naming a variable
take a walk while thinking how to optimize a function
keep your logical brain working
English
intern.🍓 retweetledi
Low-agency: “Can I do this?”
High-agency: “I’m doing this unless someone stops me.”
The high-agency person has realized that most permissions are granted retroactively. It’s easier to get forgiveness than permission because once something is already done, the default switches from “no” to “well, I guess it’s fine.”
Kpaxs@Kpaxs
Here a controversial take: most of the authority that exists in any organization was never formally granted to anyone. It was assumed, exercised, and then retroactively legitimized by the fact that it worked.
English
If you want to cook w/ @Polymarket v2 - here is the indexer for getting all the data realtime:
github.com/enviodev/polym…
English
>cough<
generating dependencies via AI
is more secure than consuming oss ones
GIF
The Hacker News@TheHackersNews
🛑 WARNING: Bitwarden CLI was compromised in a supply chain attack. @bitwarden/cli@2026.4.0 included malicious code after attackers hijacked GitHub Actions, stole secrets, and pushed a tampered version to npm. 🔗 Learn how the attack worked → thehackernews.com/2026/04/bitwar…
English
you see what they did
bought openclaw to control competition and sunset it longterm
fuck self hosted hobbists
own the apps, own the infra
OpenAI@OpenAI
Introducing workspace agents in ChatGPT—shared agents that can handle complex tasks and long-running workflows across tools and teams.
English
at this point i think the white house intern is an ex CT sol memecoin shiller
The White House@WhiteHouse
LOCKED IN MODE: ACTIVATED 🇺🇸🔥 It’s Monday. No excuses. Just wins. 🦅
English
underrated: using @fiberplane to dissect big changes into smaller chunks, embrace the exponentials, come back and check the work
Sherwood@shcallaway
OVERRATED: running tons of agents in parallel; working on too many things at once; perpetual context-switching; opening lots of low-quality PRs that may never land. UNDERRATED: using one or two agents at a time; focusing on the task in front of you; thinking deeply; finishing stuff; making your code works in prod.
English
Considering 90% of crypto frontends runs on Vercel
Super happy to have taken the contrarian route and go self hosted from the start
Vercel@vercel
We’ve identified a security incident that involved unauthorized access to certain internal Vercel systems, impacting a limited subset of customers. Please see our security bulletin: vercel.com/kb/bulletin/ve…
English
Crypto once again is in the LinkedIn era of AI
You can get SOTA uncensored models directly from huggingface by just searching for "Heretic"
You don't need TAO subnets to fine tune for modern obliteration
People out there are fine tuning uncensored sota OSS models with consumer GPUs
The information gap is so unreal here
Venice@AskVenice
Venice Uncensored 1.2 is now live. Developed with @dphnAI, this model delivers the most uncensored version of Mistral 24B. Upgraded with vision support, a 4x larger context window, and stronger tool-use capabilities. Trained on Bittensor Subnet 4 @TargonCompute.
English
