Research Papers - SiliconNoon (Page 17)

The Guardrails

Amodei's pacing plan puts outside auditors inside Anthropic

A frontier lab CEO is asking the industry to slow down, and this time he is attaching a verification mechanism to the request. Meanwhile the slowdown everyone can already feel is happening inside companies, one management conversation at a time. Anthropic CEO Dario Amodei published "We Must Pace the Frontier," an essay arguing that capability gains are outrunning the work needed to control them, and that the fix is deliberate, verifiable deceleration rather than a halt.

The Frontier

Motus2: a robot world model that grades its own actions

Shengshu Technology, the Chinese world-model lab, shipped the rare robotics release where the loop actually closed: one model that acts, imagines, and scores itself — then learns from the grade. Shengshu released Motus2, a "self-evolving" general world model for dexterous manipulation that puts three capabilities — acting, predicting, and evaluating — inside one shared-parameter video-action model, according to the project page and the paper on arXiv.

The Frontier

Google's TimesFM-3 forecasts every variable at once

A quiet-but-real model release out of Google Research, and a state-backed number for just how much computation the agent era is expected to eat. Google's TimesFM-3 became the first model in the family that can forecast many variables at once — and it fills the entire future in a single pass. The 330-million-parameter transformer chops each series into patches of 32 points, normalizes them onto one scale, then alternates two attention directions: along time within a series (past values only, so…

The Arena

Ex-Li Auto execs raise nine figures for a robot foundation model

Robot brains and AI infrastructure drew the money this hour: a Chinese embodied-model lab run by two of Li Auto's most senior AI executives closed a nine-figure angel round, and one of OpenAI's most visible former leaders resurfaced on the board of a data-center company heading for an IPO. Xieyue Intelligence has raised several hundred million yuan in an angel-plus round, announced September 12, with Linear Capital, Junshan Capital, Hongyi Capital and Yinshan Capital participating.

The Stack

OpenAI's training opt-out has a reasoning-token-shaped hole

Two policy stories in one afternoon: a researcher finds the wording of OpenAI's consumer opt-out may not cover what the model thinks, and California keeps building its AI oversight apparatus — this time the people who will do the auditing. The guarantee ChatGPT's paid users opt into may stop short of the hidden reasoning that contains everything they typed.

The Arena

OpenAI drops Caltech math hackathon after mathematicians cry 'slop'

The friction between AI and its neighbors is getting legal teeth this midday: a lab retreats from academia's loudest critics, an infrastructure startup hires a lab's alumni for its IPO run, and a Chinese court puts a prison term on jailbreaking a model. OpenAI has withdrawn its sponsorship of the Caltech Mathathon, the student-run hackathon for research-level mathematics, after an open letter from mathematicians accused AI companies of advancing a campaign of "slop mathematics." The event …

The Arena

Fields Medalist founds institute to prove AI safe — like cryptography

Two kinds of trust shipped today: a Fields Medalist building the mathematics of "safe", and two Chinese platforms engineering agents that can explore without breaking things. **Jacob Tsimerman, who won math's Fields Medal in July, is founding an institute whose whole premise is that AI safety has no shortcut yet — and he wants to build one.** The Mathematical AI Safety Institute (MAISI), reported by The New York Times, will start research in January 2027 in the San Francisco Bay Area with 10…

The Frontier

Loop a transformer twice, save 18% of training compute

Two stories from the research world today: an architecture paper that puts hard numbers on one of this year's more interesting ideas in model design, and a peer-review system that is quietly rewriting the economics of who gets to submit. Re-running a transformer's middle layers twice can cut training compute by up to 18% — if you pay for the extra passes honestly.

The Frontier

GPT-6 Astra outscores every specialist at antibody prediction

The rule that AI-for-science requires dedicated models just broke. OpenAI's GPT-6 Astra topped a blind benchmark for predicting whether antibodies can become drugs — beating purpose-built systems without any domain-specific training. GPT-6 Astra scored 37.98 on the DDD Benchmark's antibody developability module, outperforming every specialized model in a blind evaluation built by Insilico Medicine.

The Guardrails

The researchers who left say AI transparency is voluntary — Trump says he has no extinction concerns

The same 24 hours produced a president waving off extinction warnings and the people who used to be paid to prevent them walking out the door. Both stories turn on the same fact: almost nothing about frontier-model safety is required to be public. President Trump was asked directly whether AI leading to human extinction concerns him, and his answer was no.

The Stack

Ayar Labs adds $150M more as copper interconnects hit the wall

Two ends of the AI buildout moved today: the physical layer and the policy layer. An optical chipmaker pulled in another $150 million to rip copper out of AI clusters, and China's industry ministry put hard 2028 deadlines on rebuilding its software industry around AI. Ayar Labs, the startup building optical links between AI chips, has raised an additional $150 million — bringing its 2026 funding to $650 million.

The Arena

Garry Tan tells AI's accusers to relax — 'I would do nothing' about distillation

Y Combinator CEO Garry Tan says he would "do nothing" about Chinese labs distilling American frontier models — a direct counter to the loudest story in AI right now. Speaking with CNBC at YC's Demo Day, Tan argued the right goal for a regulator isn't blocking distillation but striking a balance where open-weight models keep giving people "freedom and access" while frontier models keep a price premium that funds their research.

See all