The latest AI news, in focus.

The Sequence Knowledge - Issue #959: Learning RSI: How Models Learn to Direct Their Own Updates

SEAL and the architecture of persistent adaptation

b11926

glm5-next : QSA without kq_mask (+ DSA) (#31009) * glm5-next : QSA without kq_mask * DSA: dsa top-k mask change tested with layers 0,3 of DeepSeek-V3.2 **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11926/llama-b11926-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11926/llama-b11926-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11926/llama-b11926-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11926/llama-b11926-bin-ubuntu-x64.t

Rokid has unveiled three new AI-powered wearable devices, expanding its display and interaction solutions.

On October 27, Rokid launched the AI glasses Rokid Glasses 2, the full-color AI glasses Rokid Color, and the interactive ring Rokid Ring.

AI Short Drama: The Secrets That Office Workers on the Assembly Line Can’t Help But Share

Those who dive into AI short-form dramas have no dreams—only an insatiable hunger for traffic.

Regarding the Kimi quota issue, Anthropic has found a solution.

Why isn’t your quota being used? Most of it is going toward “holding ineffective meetings” for sub‑agents.

Annual salaries of 100 million to poach scientists, and monthly salaries of 100,000 to recruit interns—robotics companies are making exorbitant offers to snap up talent.

The battle for embodied intelligence talent.

Neocloud: This wall on the verge of collapse—Jensen Huang is propping it up, while Musk is pushing it down.

The first domino in the AI bubble.

Anthropic’s Silicon Valley Past

The next round of reshuffling in Silicon Valley may well begin right here.

The HKEX AI Index is here, and Southern Fund is the first to take the stage: How significant is this move?

Index implementation plus ETF authorization have provided Hong Kong‑listed AI and semiconductor stocks with new allocation tools and a fresh source of capital.

How can developers capitalize on the foundational benefits of the AI era? The CEIC × 1024 Shenzhen event unveils the new infrastructure for AI and HarmonyOS development.

During the 2026 Changsha · 1024 Programmers’ Day event, the CEIC Consumer Electronics Innovation Conference and CSDN joined forces in Shenzhen to present a cutting-edge tech extravaganza for developers. As Yang Tao, Executive Deputy Director of the CEIC Organizing Committee, noted in an interview: “Ultimately, every experience innovation in the consumer electronics field is brought to life by the code written by our programmers.” At this pivotal moment in technological advancement, the evolution has moved from plain hand-written code to leveraging intelligent agents (Agents) and powerful on-device APs...

Data center expansion has sparked local opposition across the United States, making it a key issue in the 2026 midterm elections.

According to a report by The Verge on October 27, 2026, communities across the United States have mounted widespread opposition to the large-scale construction of data centers and the soaring costs of associated infrastructure, with this issue deeply entangled in numerous congressional and gubernatorial races. Tech capital is accelerating its entry into the fray: the super PAC “Leading the Future,” backed by Marc Andreessen and Greg Brockman, has already spent $45 million within a year, while “Public First Action,” supported by Anthropic, has also joined the effort.

Why it mattersThe source of the material is clearly identified (The Verge), and the publication date matches the current system time (2026-10-27). The content focuses on verifiable public policies and electoral developments, including specific entities, monetary amounts, timelines, and organizational relationships, thereby meeting the requirements for timeliness and factual grounding.

iFLYTEK’s Liu Qingfeng: iFLYTEK Spark X2.6, trained on fully domestically produced computing power, is betting on edge-side models to “bypass” restrictions.

Liu Qingfeng stated, “We don’t want iFLYTEK to be the only one conducting training on domestically produced computing power; by speaking out loudly, we hope more players will join in.”

Selling AI Tokens Gets Harder as the Discount Trade Unravels

Model credits are showing up in festival promotions and carrier-style packs. Thin resellers that thrived on opaque markups face quality doubts, enforcement, and better-funded rivals. Durable intermediaries will charge for routing, trust, and integration—not for a secret price cut.

Most computing power, slowest progress: ByteDance’s large model stuck in a race-to-arms dynamic.

The old “horse racing + short-cycle evaluation” approach can no longer be sustained.

b11925

ci : use ccache depend mode for ROCm release builds (#31231) **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11925/llama-b11925-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11925/llama-b11925-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11925/llama-b11925-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11925/llama-b11925-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b1192

From Large Language Models to World Models

Yan Deli, a senior expert at Tencent: LLMs are probabilistic models. They do not truly understand language; rather, they predict the next token (Next Token Prediction). Essentially, they are monotonous “word-chain” machines—moving forward one step at a time without retracing their path. At their core, LLMs are probabilistic models that learn, from statistical co-occurrence patterns, “what often appears alongside what,” and output “this is usually the case,” without grasping “why it is so.”...

After reading the prospectus, I finally understand why Anthropic no longer cares about its users.

The big one has finally arrived. On October 26, Anthropic publicly filed its S‑1 registration statement with the U.S. Securities and Exchange Commission, preparing to list on the Nasdaq Global Select Market under the ticker symbol ANT. Media reports suggest the company’s valuation is being discussed in a range of $1.5 trillion to $2 trillion, though the proposed offering price range remains undisclosed. Within 24 hours, this more than 300-page prospectus had already been scrutinized inside and outside the tech and finance communities, with everyone focusing on the most eye‑catching figures: revenue of $16.286 billion in the first half of 2026, compared to just $1.163 billion in the same period last year; operating profit of $974 million in the second quarter, marking the company’s first-ever quarterly operating profit; and a net loss of $80.7 billion, largely driven by $72.388 billion in changes to the fair value of convertible bonds, derivative liabilities, and SAFE notes—non‑cash accounting expenses. Yet beyond these Wall Street narratives, domestic developers and everyday users experience something quite different. In the eyes of mainland Chinese users, Claude is arguably the AI product currently facing the most account bans, the strictest risk controls, and the most frustrating user experience. Why does a company that has built a state‑of‑the‑art large model consistently appear so indifferent when dealing with ordinary individual users? After poring over the financial footnotes and risk disclosures in the prospectus, the answer is actually right there in the numbers. In Anthropic’s business landscape, personal users contribute only a very small share of its revenue. 01 Less Than 20% from Individual Users To understand who a company truly prioritizes, the simplest approach is to examine its revenue breakdown. In Note 2 of the financial statements attached to the prospectus, Anthropic splits its revenue into two main categories: usage‑based revenue and subscription revenue. The contrast is stark: for the full year 2025, usage‑based revenue totaled $3.796 billion, while subscription revenue was just $789 million—roughly an 83% to 17% split. By the first half of 2026, usage‑based revenue surged to $13.064 billion, with subscription revenue at $3.222 billion, still leaving usage‑based revenue accounting for about 80% and subscriptions making up roughly 20%. Even more telling details lie hidden within the composition of subscription revenue. Of that 20%, there are both individual users’ Cl

b11923

ggml : relax scale to non-contiguous views (#31146) * ggml : relax scale to non-contiguous views SCALE requires src0 to be 1D-contiguous. Relax this to dim-0 contiguity with arbitrary row strides (ggml_is_contiguous_rows), same as the unary ops in the Metal backend. - ggml: relax the assert in ggml_scale_impl - cpu: address src0 rows through their byte strides (f32, bf16) - cuda: stride-aware kernel, fastdiv row decomposition - metal: support bf16 scale (guarded unary instantiations) - sycl: keep the gate at contiguous, the kernel is flat f32 Add view variants to test_scale. Verified on CPU, Metal and CUDA. Assisted-by: dsh:llama.cpp/ggml-org/Qwen3.8-Flash-Next-GGUF (RTX 5090 32GB, DDR5 128GB) * cuda : skip the softcap fusion on non-contiguous views The fused softcap kernel reads src and w

Manus has just quietly released its 2.0 manifesto.

On October 26, Zhang Tao (online alias HideCloud), co-founder and product partner at Manus, recently published a lengthy article on his WeChat public account “Qianyun Sisi,” titled “Transforming Manus with Manus.” The piece opens with an image that left a deep impression on him. He had submitted several requirements and bug reports to an engineer; when he checked back later, he found that the engineer’s computer had multiple Codex and Manus sessions open. Clicking on any one of them, he saw that the input was precisely the original requirement he’d just posted on Slack—without a single word changed. Even more surprising to him was that this engineer was simultaneously working on over ten or twenty tasks. At Manus today, this has become the norm. The article is packed with information, but a few key points are especially worth noting. Internally, Manus has created a unified code entry point called Manus-mono, accompanied by a manual designed for AI, enabling the AI system to view the entire project landscape across repositories. With a single request, one person can complete all client-side development. The mobile team has consistently consisted of only seven members, yet they manage two products and four apps, keeping their progress largely in sync with the web version. All product managers have submitted code to Manus, and roughly half of them can independently bring their requirements to production. Some product managers even turn their written requirements into video presentations. Meanwhile, the marketing team now has a “digital colleague” that schedules its own tasks, proactively reminds everyone of project deadlines, and even reaches out across teams to clarify feature details. In September, the company’s per-capita AI spending exceeded $10,000, with a small number of individuals consuming as much as $70,000 to $80,000 in a single month. Zhang Tao dedicated an entire section to addressing external skepticism. The timing of the article’s release is also noteworthy. At the end of September, Manus launched version 2.0 along with its personal AI agent app, Cue; in early October, the parent company Butterfly Effect announced the completion of a new round of financing exceeding $500 million, while simultaneously opening 17 positions in Beijing. Seen against this backdrop, the article reads less like a routine update and more like a manifesto marking the Manus team’s entry into the 2.0 era. It discusses how the organization leverages AI to evolve, while revealing the company’s product roadmap, cost structure, and hiring standards. 01 Manus itself serves as a showcase for 2.0 If you compare the internal practices described in the article with the feature list of Manus 2.0, you’ll find…

Anthropic still has to keep “charging ahead”

The prospectus shows that in the first half of 2026, the company generated revenue of US$16.286 billion, roughly 14 times that of the same period last year, while its operating loss narrowed from US$2.135 billion to US$693 million.