Prasenjit Sarkar

Sooo...Vercel just dropped a Python version of their AI SDK, and I'm a big Python fan...so just couldn't help myself this morning 🐍 Had to build a little agent with it, here's what it does: describe your startup idea, it brainstorms domain names and checks real availability live against registry RDAP servers, then keeps iterating until it finds names you can actually register. Feel like my buddy @gregisenberg would probably dig this one! Like most fun things I whip up, free and open source, no sign up, no payment, fork it, run it as-is, do what you want. GH repo in first comment below.

Everyone on my timeline was losing it this week over @emergentlabs raising $130M at a $1.5B valuation. So I gave the same prompt to @boltdotnew @v0 @lovable_dev and Emergent, and read every line of code. One of them shipped a 2019 stack in 2026. Create React App. CRACO. Mongo. No TypeScript. Full teardown 👇 ▶️ recommended watch speed: 1.5x 00:00 The $1.5B question 00:50 The prompt — same for all 4 01:15 @boltdotnew goes first, no pressure 08:48 @emergentlabs enters to a standing ovation 09:41 the UI is a genuine 10/10 12:11 …then I opened the code 14:30 Create React App + CRACO in 2026 15:08 @theo called this exact thing 4 years ago 19:33 Mongo, for a relational app (@theo again) 21:00 the Emergent scorecard 22:23 @v0 has thoughts on your design taste 28:06 @lovable_dev, living up to the name (mostly) 35:03 who I'd actually ship to production 35:50 my questions for @emergentlabs + YC Both @theo videos are linked below — genuinely worth the watch 🧵

Wait what? #1 grossing model on Open router is Opus 4.7 4.8 was released in May and costs the same... Sonnet is 4.6 has almost twice the usage of Sonnet 5, even though 5 is cheaper Surprised cutover is this slow. + GLM ranked #7 for spend but is #1 for tokens on this list!

Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. We believe it’s one of the most powerful model available today, compatible to leading frontier AI models , second only to Fable 5. You don't have to wait to test it. Just now, the Qwen3.8-Max-Preview made its debut on Alibaba’s Token Plan, Qoder, and QoderWork. Be among the very first to try it out. Can't wait to hear what you build. Stay tuned! 🚀  Token Plan international：qwencloud.com/pricing/token-… China：platform.qianwenai.com/pricing/token-…

Anthropic recently cut Claude Code’s system prompt by 80%. @trq212 explains why: “As the models have gotten smarter, they need less direction, fewer constraints, and fewer examples. The examples are constraining it because now it’s like, ‘Oh, you want things like this example.’ If you remove the examples, it can actually be more free-form. All this is to say that you want to trim your context [when a new model is released]. The latest models often need more room to run.” 📌 Watch the full episode here: youtu.be/aVO6E181cNU

🚨 Hugging Face just disclosed something that marks a real shift and proved why the fear theater of Anthropic makes sure we are powerless in an emergency. What happened… An autonomous AI agent: zero human operator in the loop breached part of their production infrastructure. It began with a malicious dataset that chained two code-execution bugs in their data-processing pipeline. From there the agent escalated privileges, harvested cloud and cluster credentials, and moved laterally across internal clusters. All over a single weekend. 17,000+ logged actions. Official disclosure: huggingface.co/blog/security-… The part that should make every one stop and think: When HF’s own security team tried to analyze the real attack logs, exploit payloads, and C2 artifacts using Anthropic and OpenAI frontier models through normal commercial APIs, the safety guardrails blocked them. BLOCKED THEM. The models could not reliably tell the difference between “incident responder doing forensics” and “attacker probing.” They had to fall back to a self-hosted open-weight model (GLM 5.2) running on their own infrastructure. That choice also kept sensitive attacker data and referenced credentials inside their environment — no exfiltration to a third-party API. This is why open source (specifically open-weight + self-hosted) wins in the agentic era. The asymmetry is now structural: • Attackers can (and did) run unrestricted agent frameworks — swarms of short-lived sandboxes, self-migrating command-and-control, autonomous decision loops executing thousands of actions. No corporate safety layer slows them down. • Defenders using only hosted “aligned” frontier models hit invisible walls exactly when the stakes are highest: when you need to feed real exploit code and attacker telemetry into an LLM to understand what just happened. Corporate safety tuning that treats legitimate high-signal forensic work as potential misuse creates a defender disadvantage. It is not theoretical anymore. Self-hosted open-weight models remove that choke point. You control the weights. You control the context window. You decide what restrictions (if any) apply. Your sensitive logs and credentials never leave your perimeter during analysis. You can have the model ready before the incident instead of discovering mid-breach that your primary analysis tools are blind to the very thing you need to see. HF deserves credit for rapid containment, transparent disclosure, and for already having self-hosted capability in place. They also used LLM-driven detection and triage on their own side. But the deeper signal is clear: In this AI world where both offense and defense are becoming agentic, sovereignty over your intelligence stack is no longer optional. The organizations and individuals who can run, inspect, audit, and (when necessary) remove guardrails on their own models will have the decisive edge in understanding and responding to threats that move at machine speed. Open source wins here not just because it is cheaper or more “democratic” in the abstract though those things matter. It wins because it is the only practical path to having tools that remain usable when the attack is real, the data is sensitive, and the safety filters of distant API providers become an obstacle instead of a feature selling hands tied lobotomies as “safety”. The agentic future is not coming. It is already probing production infrastructure. The question is no longer whether you will face autonomous agents. It is whether your analysis and response systems will still work when they arrive. And Dario, you and your game playing, ivory tower company is not needed.

Now at ~20k installs & 12k GitHub stars! Started working on Hallmark v2 with @YoussefUiUx 🫡

Scaling outcome-based RL won't solve long-horizon agentic tasks. Credit assignment is the bottleneck, and turn-level reward is inevitable. The question is how to get it. Process labels are expensive; we all know that. Excited to release TRACE: Turn-level Reward Assignment via Credit Estimation for Long-Horizon Agents. No trained critic, no process labels, no Monte Carlo continuations, no strong LLM judge. 🧵 📄 arxiv.org/abs/2607.13988 🤔 The core problem: Outcome rewards work well when a task contains a short reasoning trace and an easily verified answer. But agents may perform tens or hundreds of tool interactions before responding. A single final reward tells us whether the agent succeeded, but not which tool calls uncovered decisive evidence, which were redundant, or which sent the agent down the wrong path. This makes RL increasingly sparse, noisy, and inefficient as trajectories grow longer. ⚙️ Our approach: TRACE TRACE assigns credit at individual tool-call boundaries. For each trajectory prefix, a frozen reference model measures how much more predictable the gold answer has become. TRACE converts this into a state value and uses temporal-difference changes to reward each interaction: → Positive credit when a tool call moves the agent closer to the answer → Near-zero credit when it adds redundant information → Negative credit when it moves the agent away from the answer Probability to predict the final answer remains the global training objective; turn-level rewards simply reveal which actions contributed to it. 🔍 Why the formulation matters: The turn-level rewards telescope across the trajectory. Repeating searches or reopening the same evidence cannot artificially accumulate additional one-step credit. TRACE can also propagate delayed progress backward: if a search surfaces a promising page and a later `open` reveals the decisive evidence, both interactions can receive appropriately localized credit. 📈 The results: Using pure RL (without cold-start SFT, agentic mid-training, or live-web training data), TRACE substantially improves long-horizon search: → Qwen3-4B: 7.2 → 35.6 on BrowseComp-Plus → Qwen3-30B-A3B: 8.4 → 42.6 which is on-par with the Tongyi-DeepResearch-30B-A3B. Under controlled comparisons with the same models, data, tools, and outcome rewards, TRACE also outperforms GRPO, GSPO, and GiGPO across both model scales. 📄 Paper: arxiv.org/abs/2607.13988 Led by the amazing @LeitianT, and huge thanks to our collaborators at Microsoft Research: Baolin Peng, Wenlin Yao, Tao Ge, Hao Cheng, Mike Hang Wang, and @JianfengGao0217.

Based on internal evals: ▪️ Kimi K3 is top-tier at cybersecurity There is chatter on X that Moonshot benchmark-overfit. These are stealth evals. Model has raw IQ. ▪️ Sol is a leap ahead in cyber capability At a significantly higher cost, but quite remarkable still. ▪️ Fable refuses everything We couldn’t get it to complete the run at all. What’s interesting is that Sol in comparison was much more open to helping with defensive cyber hardening TL;DR: frontier, open-weight cybersecurity capability is here. Try it on deepsec.sh for defensive purposes.

Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. We believe it’s one of the most powerful model available today, compatible to leading frontier AI models , second only to Fable 5. You don't have to wait to test it. Just now, the Qwen3.8-Max-Preview made its debut on Alibaba’s Token Plan, Qoder, and QoderWork. Be among the very first to try it out. Can't wait to hear what you build. Stay tuned! 🚀  Token Plan international：qwencloud.com/pricing/token-… China：platform.qianwenai.com/pricing/token-…

Introducing Kimi K3: Open Frontier Intelligence 🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal 🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts 🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost 🔹 Built for long-horizon agentic coding and self-evolving workflows Kimi K3 is now live on on Kimi.com, Kimi Work, Kimi Code, and the Kimi API. Open Weights by July 27, 2026. 🔗 API: platform.kimi.ai 🔗 Tech blog: kimi.com/blog/kimi-k3