Starbase Louisiana, Jalapeño, Claude 메모리와 로컬 에이전트 실행

Introducing Starbase, Louisiana – a high-cadence launch site that will enable Starship to open a pathway to the stars for humanity. Thank you to @LAGovJeffLandry and the great people of Louisiana for all of their support as we work together on this audacious project → http:// spacex.com/sites/starbase -la …

Elon Musk

@elonmusk

13시간 전

Starbase Louisiana will ultimately have over a dozen launch towers, enabling more than 30 Starship flights per day and making it the biggest launch site on Earth! SpaceX makes sci-fi real.

Since announcing Jalapeño, our first custom inference chip, we’ve been testing it and the system around it. The results show a major advance: more intelligence from every watt and faster responses, delivering both higher throughput and lower latency in one architecture without sacrificing efficiency.

We plan to begin deploying Jalapeño in OpenAI’s compute infrastructure by year-end. It’s the first step in a multigenerational roadmap: Gen 2 is deep in development, and Gen 3 is taking shape. Each generation will push efficiency and speed further.

Claude

@claudeai

12시간 전

Claude now has one memory across chat and Claude Cowork, and you decide what's in it. Hand Cowork a task and it starts from what Claude already knows from your chats: the project you talked through, your manager's preferences, or the client from last quarter.

Claude

@claudeai

12시간 전

Everything Claude remembers is saved as a list of topics in Settings, where you can read, edit, or delete each one. Memory also updates on its own as you chat, saving new details as they come up. You can also say "remember this" to save something specific.

Today we’re launching Portable Computer on @NVIDIA DGX Spark. Portable Computer is a fully local version of Perplexity Computer, where the entire runtime: orchestrator LLM, subagent LLM, agent harness all run on your local hardware. No cloud dependency.

You can now fine-tune Qwen3.8-27B for free with our notebook! Local training works on 24GB VRAM. Unsloth trains Qwen3.8-27B 1.5x faster with 50% less VRAM. GitHub: https:// github.com/unslothai/unsl oth … Qwen3.8-27B Notebooks + Guide: https:// unsloth.ai/docs/models/qw en3.8/train …

Viv

@Vtrivedy10

14시간 전

this is a practical guide (+ updated skill) on how we use real world data like traces + human feedback to make synthetic environments + evals so we can measure, harness engineer, and post-train our agents a few main components: - World Knowledge gathering of the systems our agent will interact with - Using Specs to coordinate what will be built —> World Spec & Task Spec - Human-Agent collaboration to iteratively edit a Spec and infuse human feedback - An execution pipeline that takes an agreed on Task Spec and generates an environment + task - Fixing design flaws in the task and environment by actually running agents in the environment & mining the verifier scoring + traces @harrison_chase , @nick_hollon and I spent a ton of time going back and forth on design decisions and the overall flow - how much a coding agent should prompt for human feedback - what specs do we need and what do they contain - how to create and update world knowledge this new flow is packaged in our updated eval-engineering skill. we want to help every team own their intelligence which means owning the pipeline to convert their valuable data into Tasks+Environments that improve their agent over time check out the skill and reach out if you’re thinking about this! https:// skills.sh/langchain-ai/l angchain-skills/eval-engineering …

Nick Hollon

@nickhollon10

14시간 전

@Vtrivedy10 and I have spent a lot of time thinking about how to efficiently build evals and environments. our eval-engineering skill detailed here helps spin the environment generation loop very quickly!

Impressive work from Alibaba. (bookmark it) If you build long-running agents and keep rewriting your memory schema, take a look at this approach. It basically treats agent context management as a programming task. Here is how it works: It backs each agent session with an append-only event log and a sandboxed, persistent Python kernel. Tool outputs, retrieved history, and derived state bind to typed variables across model calls instead of being serialized into the prompt every turn. Model-written code searches and transforms that state, and only explicitly printed projections enter the working view. The event log keeps lossless ground truth, so nothing has to be committed to a compressed form before you know what will matter later. When the working view nears its budget, stale spans are evicted but stay recoverable. An eviction index keeps compact landmarks tied to exact event-log addresses, so the agent navigates straight back to a region instead of searching the whole log. Results: with Qwen3.8-Max, 94.8% on LongMemEval_S, 73.1% on BEAM_10M (5.1 points over the best published memory system), and 86.7% on LOCA_256K. Treating context management as a programming task means it inherits every future improvement in model coding ability. Paper: https:// arxiv.org/abs/2608.21690 Track more trending AI papers in our academy: https:// academy.dair.ai

Introducing S1, our new foundation model that learns from one example. It can be taught 10-minute long tasks that it has never seen before, from one video prompt without any fine-tuning. Watch S1 operate in real-time via in-context learning:

Introducing Composer, a section-by-section song editor. Start with your own lyrics or track, a blank page, or a prompt, then shape the track piece by piece until it’s finished.

Mati Staniszewski

@mati

15시간 전

Stripe is now an ElevenLabs customer! When Piotr and I started @ElevenLabs , one of the first things we did was set up a @Stripe account and every subscription we’ve collected has run through their infrastructure. We're excited that Stripe is now using ElevenAgents to interact with the businesses it serves and ElevenCreative to support its brand and marketing work. Stripe built the payments layer for the internet, serving millions of businesses around the world. Our goal is to make ElevenLabs that layer for customer interaction.