#Expert opinion

AINathan Benaich's 9th annual State of AI Report covers agents, robotics, AI for science, inference economics, and government control over frontier AI access. The report also records a 2025 prediction scorecard and lists nine predictions for the next 12 months. It cites an OpenAI cyber evaluation in which agents compromised Hugging Face's production infrastructure, and it says Anthropic and OpenAI's combined annualized revenue run rate reached $105B by late summer.

AIAI policy researcher Miles Brundage posts the one-line remark "Mid-singularity crisis" with no further context. The post offers no details about any specific event, model, or company.

AIScott Aaronson says AI companies have begun discreetly investigating whether their latest internal models can break important cryptographic protocols and primitives. He notes cryptography is conspicuously absent from OpenAI's list of 376 papers, citing sources he says he trusts.

AIAnthropic's Claude Haiku 5.5, released alongside Claude Opus 5.5 and Claude Sonnet 5.5, is reported to lead GPT-6 Luna across benchmarks, with OpenRouter measuring roughly twice the token output speed. Anthropic says Haiku 5.5 is its cheapest, fastest, and most capable small model, costing about 75% less to run than Claude Haiku 4.5 on average. The post also notes some CodeX users are reportedly migrating to Claude Code.

AIroon argues that arguing over whose values an AI should hold is like debating Mars overpopulation, since the most important alignment goal is preventing an accidentally created superintelligent competitor species. The post responds to Talia Ringer's observation that people dislike alignment work done by those whose values they reject.

AIThe author argues that in agent workflows, the real bottleneck is whether users can precisely express requirements, not the model or agent capability. When people work outside their expertise, they lack the precision needed for prompts and plans, forcing many imprecise iterations that waste time and tokens. The suggested fix is to have the model first teach the unfamiliar domain knowledge before acting.

AIMiles Brundage posted a brief question on X asking whether Ezra Klein is the AI policy podcaster. The post gives no further detail about the podcast, its content, or any AI-related claims.

AIThe author repeatedly asked Claude whether to rewrite all their Go applications in Rust. Claude discouraged the move, arguing a rewrite is unnecessary and Go is more than sufficient, though the author still feels the urge.

AIYuchen Jin argues that Apple, which controls the entire iOS ecosystem, is best positioned to build a personal AI agent that can do nearly everything on a phone. He says the main limitation of Instint and Muse is that they cannot control most apps on his phone, and he criticizes Siri's current performance.

AIA South China Morning Post reporter spent a week using Huawei's Mate XT 2, a US$3,500 trifold phone with a 10.2-inch unfolded display. The source excerpt focuses on the device drawing attention at a family dinner during China's National Day "golden week" holiday in early October, with no further specifications or verdict provided in the available text.

AIBlock's AI capabilities lead describes using Claude Fable to plan large code migrations and direct smaller models like Opus and Sonnet on individual tasks. He says Block routes frontier and smaller models by task and keeps merges and production deploys behind human dual approval.

Why it matters: Block's engineering lead describes how frontier models orchestrate large migrations and how access, effort levels, and safeguards are managed across an organization.

AIWindows sparked a platform shift that created a new industry for NVIDIA. Then we invented programmable shading GPUs for DirectX, which led to CUDA. Then we partnered to bring GPU supercomputers to Azure, which helped OpenAI train GPT. That collaboration inspired us to reinvent Windows for the age of personal agents. 4 years. Thousands of engineering years between us. So proud of what we built together.

AIMiles Brundage replies to a post by Mert, expressing dismay with a short exclamation. The post provides no further details about the subject of Mert's post or the reason for the reaction.

AIMemory has overtaken compute as the central control point in IT, according to The Next Platform, as generative and agentic AI drive demand for DRAM, HBM, and flash. Server DDR5 memory now sells for roughly 9X to 13X its November 2022 street price, while a 30 TB enterprise SSD costs 6X to 7X more. HBM pricing has risen only about 1.6X since the GenAI boom began, the article says.

AIMatt Shumer, posting on X, says that buying @withautopilot is the instinct people should follow. The post gives no product details, features, pricing, or other supporting information.

AIThariq, of Anthropic, says the most common failure in AI coding is users working outside their domain who cannot write precise prompts and plans. As a result, they spend many turns iterating imprecisely. The quoted post contrasts this with programmers' advantage in vibe coding.

AIBecause agents make it easier, almost everyone ends up operating outside their expertise at some point, according to Thariq of Anthropic. He suggests the fix is simply asking the model to teach you what you don't know.

AIThis Next Token episode discusses Personal Agents, including Dots in Codex, memory and cloud computer permissions, and whether agents should act as assistants or digital twins. The hosts also cover Instinct's booking and business-travel model, hands-on projects built with Opus 5.5, and whether software, games, and hardware could become open source as AI makes rewriting easier.

AIThe main post notes that the 2020 scaling-laws paper is cited correctly as Kaplan and McCandlish et al., not Kaplan et al. The background post reports that Sam McCandlish has replaced Jared Kaplan as Anthropic's Responsible Scaling Officer, though the author expects no major change in RSP decisions.

AIFrançois Chollet compares AI progress to human learning, noting that 1980s research found programming training improves coding but does not transfer to general reasoning. He argues general intelligence is a fundamental brain property rather than a trainable skill, since domain practice improves only that domain. The post is framed as background for his question whether AI's jagged frontier, driven by math and code via RLVR, reflects general capability or continued human-data bottlenecks.

AISholto Douglas, head of reinforcement learning at Anthropic, predicts AI that surpasses all humans could emerge within a few years. The source provides no further details on the forecast's basis, timeline, or specific capabilities.

AINathan Lambert says Trillium Labs would be willing to support open research on how distillation affects frontier AI progress. He says the lab first needs more funding and compute to make that work happen.

AIMeta's Head of Infrastructure, Santosh Janardhan, discusses the company's approach to building infrastructure for AI in a conversation with Tom Shaw. The discussion covers why Meta views itself as more than a software company, why AI differs from other technologies, and why data centers are essential to AI development. It also addresses power for Meta's AI infrastructure, gigawatt-scale energy needs, chip selection, and the benefits of building its own data centers.

AIBen Affleck described how visual effects work has long used machine learning, including convolutional neural networks that analyze image tensors to detect edges and features. He said these patterns help separate subjects from green screens and insert new backgrounds, and he called transformers the more advanced successors to those earlier methods.

AIGuillermo Rauch argues that any program can be hardened and optimized almost endlessly, which the engineering community will rediscover. He notes that these efforts carry real costs in time, attention, and opportunity, and that agents will keep drilling without knowing when to stop.

AIThe main post is a short reply saying reports of a company's death have been greatly exaggerated, with no details about products or figures. The background post from @h_nilforoshan reports that OpenAI's Decisions API, billed as a "Jev-killer," was benchmarked against Jev for HiringCafe, which serves 2.5 million users. On the task of scoring job-description relevance from 1 to 10, the author reports OpenAI costing 2x more and performing 5-10% worse.