Sabitlenmiş Tweet
Anuranjan
1K posts
Anuranjan
@anuranjan
Chief Truth Engineer | Building @AskKaval & @RealworkApp
👉 Katılım Şubat 2019
723 Takip Edilen138 Takipçiler
A scam takes one tap to forward. Verifying it takes 47 background operations. And you have exactly 4 seconds to execute them before the user clicks the link anyway.
The hardest part of AI fraud prevention is the latency constraint. Every message sent to Kaval triggers this graph:
> 3-layer routing (handles 80% of intent deterministically, zero LLM cost)
> Parallel specialist dispatch
> Evidence-grounded verdicts (policy > LLM opinion)
Under 10 seconds.
3 months ago, this was one Python script. Today, it’s 71k lines of code and 1,107 tests. It’s far from perfect, and we have a long way to go, but it's a start.
English
@himanshustwts How cool would it be to have such classes and clg culture here 💔
English
technology wont make universities obsolete but actually it will force the best ones to evolve.
what is happening at stanford might be the first instance of that like sharing coherence with frontier on the levels of chips + models + infra + products + policy under one roof.
Anjney Midha@AnjneyMidha
We have now updated the @CS153Systems site with our second batch of speakers including @LisaSu @satyanadella @garrytan @sdianahu @matiii and @andi_blatt Next week we'll unveil some surprises that are big wins for students and the broader ecosystem
English
We’re just getting started 🙌
Sarvam for Developers@SarvamForDevs
@anuranjan @SarvamAI Hey @anuranjan 10,000 credits added to your account (anuranjan.vikas31@gmail.com). Great work on Kaval.chat — protecting families from WhatsApp scams with voice notes in native languages is a super valuable project. Excited to support it
English
@SarvamForDevs @SarvamAI You guys are the best! Time to put this to full use 🙌
Thank you
English
@anuranjan @SarvamAI Hey @anuranjan
10,000 credits added to your account (anuranjan.vikas31@gmail.com).
Great work on Kaval.chat — protecting families from WhatsApp scams with voice notes in native languages is a super valuable project. Excited to support it
English
This sounds like an April Fool’s, but it’s not.
If you’re building with @SarvamAI, reply with:
• The email you’ve used to log in to dashboard.sarvam.ai
• A short note on your use case
We’ll add ₹10,000 in credits before April 1st ends
English
claude: 1m tweets about all their users eating rate limits in 5mins, third time this happened in the last 4 months, zero resets
codex: tibo sees a number on a dashboard that might be suspect, and out of an abundance of caution, hits the button. it's wednesday. thanks tibo.
Tibo@thsottiaux
Our Codex dashboards are showing increased rate of users hitting rate limits and since we don't fully understand why I have made the cautious decision of resetting the usage limits for all plans. Enjoy. I also wanted to celebrate us finding a pocket of fraudulent accounts that we banned and have helped us regain some compute. The fight against abuse never stops, but it's important to mark the moment and make it a little shared victory.
English
Our Codex dashboards are showing increased rate of users hitting rate limits and since we don't fully understand why I have made the cautious decision of resetting the usage limits for all plans. Enjoy.
I also wanted to celebrate us finding a pocket of fraudulent accounts that we banned and have helped us regain some compute. The fight against abuse never stops, but it's important to mark the moment and make it a little shared victory.
English
Shopify saved 99% by going from GPT-5 -> Qwen 3.5
Kevin Madura@kmad
$5.5m to $73k per year (!!) by: 1) decomposing business logic 2) modeling intent using DSPy 3) optimizing a smaller model to improve cost profile *while maintaining performance* Why wouldn’t this be the default pattern for folks embedding AI into their pipelines?
English
The urge to cancel Claude Max is increasing day by day
Tibo@thsottiaux
Our Codex dashboards are showing increased rate of users hitting rate limits and since we don't fully understand why I have made the cautious decision of resetting the usage limits for all plans. Enjoy. I also wanted to celebrate us finding a pocket of fraudulent accounts that we banned and have helped us regain some compute. The fight against abuse never stops, but it's important to mark the moment and make it a little shared victory.
English
i have NEVER been able to work from a coffee shop, beach or even office
signüll@signulll
i tried working from a coffee shop today instead of my office & it was impossible to get anything done. i was constantly distracted by one thing or another, from sounds to ppl moving around etc. it’s unclear to me how anyone is able to be productive working from a coffee shop.
English
@himanshustwts This is gold, I’m going to try adapt this into my agent harness…let’s see how it turns out.
English
Based on everything explored in the source code, here's the full technical recipe behind Claude Code's memory architecture:
[shared by claude code]
Claude Code’s memory system is actually insanely well-designed. It isn't like “store everything” but constrained, structured and self-healing memory.
The architecture is doing a few very non-obvious things:
> Memory = index, not storage
+ MEMORY.md is always loaded, but it’s just pointers (~150 chars/line)
+ actual knowledge lives outside, fetched only when needed
> 3-layer design (bandwidth aware)
+ index (always)
+ topic files (on-demand)
+ transcripts (never read, only grep’d)
> Strict write discipline
+ write to file → then update index
+ never dump content into the index
+ prevents entropy / context pollution
> Background “memory rewriting” (autoDream)
+ merges, dedupes, removes contradictions
+ converts vague → absolute
+ aggressively prunes
+ memory is continuously edited, not appended
> Staleness is first-class
+ if memory ≠ reality → memory is wrong
+ code-derived facts are never stored
+ index is forcibly truncated
> Isolation matters
+ consolidation runs in a forked subagent
+ limited tools → prevents corruption of main context
> Retrieval is skeptical, not blind
+ memory is a hint, not truth
+ model must verify before using
> What they don’t store is the real insight
+ no debugging logs, no code structure, no PR history
+ if it’s derivable, don’t persist it
English
@haveibeenpwned guys when are y’all indexing this…need it in the API
Archie Sengupta@archiexzzz
very big breach
English
@averycode Can it bypass blockers from all big sites? I am searching for a reliable solution for my AI agent
English
Claude fix the exploit, make no mistakes
Feross@feross
🚨 CRITICAL: Active supply chain attack on axios -- one of npm's most depended-on packages. The latest axios@1.14.1 now pulls in plain-crypto-js@4.2.1, a package that did not exist before today. This is a live compromise. This is textbook supply chain installer malware. axios has 100M+ weekly downloads. Every npm install pulling the latest version is potentially compromised right now. Socket AI analysis confirms this is malware. plain-crypto-js is an obfuscated dropper/loader that: • Deobfuscates embedded payloads and operational strings at runtime • Dynamically loads fs, os, and execSync to evade static analysis • Executes decoded shell commands • Stages and copies payload files into OS temp and Windows ProgramData directories • Deletes and renames artifacts post-execution to destroy forensic evidence If you use axios, pin your version immediately and audit your lockfiles. Do not upgrade.
English
i’m looking for ambitious founders across asia (SEA included) to come and work at our zerobase hq located in seoul for 1 month.
-4 weeks of building
-1 demo day
-get funded up to $250k
-referred to a16z speedrun
-unlimited kbbq & soju
flights included.
comment if you’re interested in coming to seoul!
English