It’s Monday, August 25th: DeepSeek 3.1 lands with bigger context and sharper reasoning, USAi launches a secure federal AI testbed, Tokyo hosts the first Genspark × AI Collective kickoff, and don’t miss our virtual, hands-on MCP workshop Wednesday,
Discover what’s happening at AI Collective chapters worldwide with our new Events Portal. Search by city, date, or format, and join thousands of builders at events across 25+ chapters and on every continent.
🌁 Based in SF? Check out SF IRL, MLOps SF, GenerativeAISF, or Cerebral Valley’s spreadsheet for more!
AI News Roundup
What you need to know: weekly by AJ Green 🤝
DeepSeek Is Back: V3.1 Out Today
News: DeepSeek just released version 3.1 of its open-source LLM with massive upgrades across context window, reasoning, multilingual capabilities, and compute efficiency.
Details:
DeepSeek now supports 128,000 tokens, enabling analysis of entire codebases or research papers without loss of context. Full demo available here.
43% increase in multi-step accuracy, especially in math, science, and agentic task workflows; supports hybrid "thinking" and "non-thinking" modes.
Supports over 100 languages with improved performance in Asian/low-resource languages and multi-modal comprehension.
Built on a 560B parameter MoE transformer using FP8 mixed-precision training; leverages multi-token prediction and advanced attention for faster, more efficient outputs.
Features 38% fewer hallucinations and outputs more structured tables, lists, and data formats.
Why it matters: With industry-leading benchmark scores like 85.5% on AlpacaEval 2.0, 94.3% on MATH-500, and Codeforces programming parity with Claude Sonnet 3.5, DeepSeek offers top-tier performance at roughly 1/68th the cost of proprietary models. Its 128K token context window enables comprehension of 400-page documents, while open-source accessibility empowers startups, researchers, and enterprises alike. The model’s rapid adoption and expanding use cases — from frontend dev automation to customer support and coding copilots — mark it as a cornerstone of the global AI landscape for 2025 and beyond.
🇺🇸 USAi launches across federal agencies
The News: The U.S. General Services Administration has launched USAi, a secure generative AI evaluation suite built for federal agencies to safely test, compare, and adopt AI at scale.
Details:
Provides access to AI tools like chat-based models, code generation, and document summarization.
Runs in a cloud-based, FedRAMP-compliant environment with models from Anthropic, Google, Meta, and OpenAI.
Includes bias evaluation, guardrails tied to the 2025 Executive Order, and workforce upskilling dashboards.
Serves as a shared services platform to reduce duplication and accelerate modernization.
Why it matters: USAi positions the U.S. government to adopt AI faster, more securely, and at lower cost, advancing the White House’s AI Action Plan. By standardizing evaluation and providing trusted infrastructure, USAi helps agencies modernize, improve efficiency, and deliver better services to the public.
🇯🇵 Tokyo Kickoff: Genspark x The AI Collective
This Friday marks the first-ever Genspark x AI Collective meetup in Japan! Expect live demos, direct access to Genspark leadership including COO Wen Sang, and an intimate gathering of Tokyo’s AI community. We’ll be looking at some of the most exciting product launches of the week:
AI Meeting Notes on Apple Watch — the world’s first, for both in-person and online meetings
Genspark AI Developer — a no-code, full-stack software app platform
Genspark AI Designer — released this past week!
Spots are limited — don’t miss your chance to be in the room!
🛠️ Collective Intelligence: Hands-On MCP (Virtual)
Turn your OpenAPI into a live MCP server in minutes this Wednesday. Learn how to generate your first-party MCP server, validate tools in Claude, and publish to an MCP registry with Emmanuel Paraskakis so your API stays trusted and tamper-free.
You’ll learn how to:
Generate an MCP server from your OpenAPI with no coding
Test MCP tools locally on Claude for fast feedback
Publish to an MCP registry and expand your API’s reach
Seats are approval-only. Secure your spot now.
Our Premier Partner: Roam
Roam is redefining remote work with a virtual workspace that mirrors the feel of a real office with co-working zones, private spaces and meeting rooms with built-in chat, video, audio, whiteboards, plus AI tools like meeting summaries and noise cancellation. Roam is supporting the AI Collective’s mission to connect builders, researchers, and leaders shaping the future of AI, reflecting their shared focus on thoughtful, human-centered collaboration.
Explore Roam and transform the way your team works together!
Are you subscribed across our platforms? Make sure to keep in touch with us online!
💬 Slack: AI Collective
𝕏 Twitter / X: @_AI_Collective
🧑💼 LinkedIn: The AI Collective
📸 Instagram: @_AI_Collective
We are a community of volunteers. All proceeds solely fund future efforts for the benefit of this incredible community.
About Noah Frank
Noah is a founder, innovation strategist, policy researcher, and community builder at the edge of AI. His work blends governance, product strategy, and revenue design to keep transformative tech aligned with human progress.
About AJ Green
AJ Green is a founder, writer, VC scout, chairman, and respected community leader in the AI and startup space. A former athlete turned tech entrepreneur, AJ is on a mission to make AI the great equalizer scaling startups, connecting ecosystems, and turning disruption into opportunity.