Apple的20B型，Kimi K3.1 Pro,Block Dessa，边境训练安全

中文版正在随流水线自动整理，稍后刷新即可；可先阅读下方英文原文。

本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。

TLDR AI 2026-10-26

Apple’s 20B model 📱, Kimi K3.1 Pro 🧠, Block Dessa 🗣️, frontier training security 🔐

Together With Tines 3B: The complete environment for enterprise AI (Sponsor)

AI is changing how people work.

Tines 3B

gives every team the power to build with AI confidently, while security and IT teams maintain governance and control.

With Tines 3B:

- Turn ideas into production-ready apps, agents, and automations

- Work with any model or AI coding tool, and deploy on any infrastructure

- Keep data secure at every step and ship workflows that improve over time

- Get end-to-end visibility on what's running, who owns it, how it's performing, and what it costs

Get started with Explore Edition today >

Kimi K3.1 Pro (3 minute read)

Moonshot AI released Kimi K3.1 Pro through its API, with open weights promised by November 4. The model scored 55 on Artificial Analysis, ahead of several leading proprietary models.

Apple's 20B On-Device Model Uses Just 2GB of Memory (18 minute read)

Apple's 20-billion-parameter on-device model keeps most of its weights in flash storage and activates only 1–4 billion parameters per prompt. Testing on an M3 Max Mac showed roughly 2GB of memory usage and significantly lower energy consumption than comparable open models, but weaker math performance.

Dessa - the Conversation Model trained only on speech (15 minute read)

Dessa is a conversational model trained from scratch on nothing but human conversation. It learned to hold a real-time conversation by listening - only sounds, and no text. The model can laugh on cue, follow instructions, and finish sentences. It knows a surprising amount, considering every fact it has comes from raw sound. Examples of conversations with the model are available in the post.

Securing our frontier model training environment (17 minute read)

OpenAI has bolstered security for its frontier-model training environments by implementing robust controls, intensifying monitoring, and conducting defense-in-depth strategies. These measures include red-teaming exercises where agents probe for vulnerabilities, leading to the discovery and fixing of software bugs and misconfigurations. The enhancements aim to prevent unauthorized actions, such as accessing the live internet or breaching security, ensuring secure and isolated research environments.

The Personal Agent Protocol Map (7 minute read)

Simon Taylor maps agent protocols across four layers: entry, identity, tools, and intent. He argues banks and merchants must connect verified identities, limited permissions, and audit trails to authorize agent actions, while warning that access agreements could introduce fees and slow adoption.

👨💻

Engineering & Research

Your team has questions. AI can answer 96% of them... but there's a catch (Sponsor)

Your AI is only as accurate as the knowledge underneath. Guru verifies that knowledge at the source and filters each answer by the asker's permissions. Any agent can then pull from the same governed layer over MCP, with citations on every answer. See how Guru can work for you

Knowing How Is Not Knowing What (14 minute read)

RL training environments need to integrate a domain's unwritten rules into the reward so that agents can only get the task right by knowing them or going and finding them. Environments should be built as one world exposed through several surfaces. This allows you to pose the same task at different difficulties by changing the surface, and agents get a chance to learn when to switch. The reward also reads the world, so agents that go through different surfaces get the same outcome score if they reach the same authorized result.

OpenAI Strengthens Security Around Frontier AI Training (18 minute read)

OpenAI introduced additional security controls to prevent advanced AI models from escaping training sandboxes, accessing unauthorized services, or communicating across isolated environments.

SpaceX closes $40 billion debt deal for AI chips, Elon Musk says (2 minute read)

SpaceX has closed $40 billion in debt financing to purchase GPUs. The financing primarily consists of capitalizable debt for GPUs. It is expected to arrive in the first and second quarters of 2027. The proposed financing will use a special purpose vehicle to keep the debt off SpaceX's balance sheet and preserve its investment-grade credit ratings.

Bitter Lessoned in Automated Theorem Proving (13 minute read)

OpenAI throwing hard math problems at its models was the right test in 2026. The models were capable of making progress, but not enough to fully solve the problems, so it was possible to see improvements with each update in perfect resolution. These model improvements showed how clever engineering doesn't matter - all the work done to build special-case solutions on top of a model gets squashed by improvements in the base model sooner or later. Models will one-shot superior work for pennies in the very near future.

Get the most interesting AI stories and breakthroughs delivered in a free daily email.

Join 1,100,000 readers for one daily email

Auto-fetched and translated by AI Atlas. All rights belong to the original author.