metir Blog
Stories, insights, and perspectives on artificial intelligence, productivity, and the future of work.
Google DeepMind
Cryptography
AI Security
DeepMind Says AI Agents Found Two Weaknesses in Legacy 1024-Bit RSA: What Is Confirmed and What Is Not
On November 6, 2026 Google DeepMind said human-guided AI models found two weaknesses in legacy 1024-bit RSA. We separate what was stated, what was left open, and how it compares with a separate September academic attack.
Nov 10, 20268 min read
Claude Fable 5.2
Anthropic AI
AI Model Selection
Claude Fable 5.2 Is Now on Metir: When to Pick It (2026)
Claude Fable 5.2 replaces 5.1 in the Metir model picker. What changes for you, when to choose it over Opus 5.5 or Sonnet 5.5, and the API changes behind the upgrade.
Nov 10, 20266 min read
NVIDIA
Stock Market
AI Infrastructure
Nasdaq's Worst Day in 19 Months: What Drove the November 5 AI Chip Selloff
Nvidia fell 9.3% and the Nasdaq dropped 5.5% on November 5, 2026. We break down the causes cited, the neocloud link, and what Nvidia's November 17 earnings will test.
Nov 8, 20268 min read
Microsoft
Neoclouds
Nebius
Microsoft Tightens Oversight of Neocloud Projects: Why Nebius and IREN Fell 16%
Bloomberg reports Microsoft is scrutinizing data center projects from Nebius, IREN and Nscale. What was reported, what the share moves mean, and how neocloud economics depend on anchor tenants.
Nov 8, 20267 min read
Data Centers
AI Infrastructure
Policy
Voters vs. Data Centers: What the November 3 Ballot Measures Mean for AI Buildout
Voters backed restrictions in nearly every local data center measure on November 3, 2026. A look at the numbers, Ohio's role, the polling shift, and why siting is becoming AI's binding constraint.
Nov 8, 20268 min read
Security
Xen
Anthropic
Xen XSA-523: Two Hypervisor Flaws Credited to Claude and What the Advisory Says About AI-Assisted Bug Finding
Xen Security Advisory XSA-523 credits Anthropic using Claude with two flaws that let a guest read and write other domains' memory, including Dom0. Versions, fixes and analysis.
Nov 5, 20267 min read
TeraWulf
Anthropic
Google
TeraWulf's Google Credit Backstop: How a $19 Billion Anthropic Lease in Kentucky Gets Financed
TeraWulf disclosed a Google credit backstop supporting debt financing tied to about $19 billion of Anthropic lease commitments at its Kentucky campus. What the 8-K says and why it matters.
Nov 5, 20268 min read
Data Centers
Ohio
Midterms
Ohio's 21 Data Center Ballot Measures and the Post-Election Stock Slide: Reading the Local Politics of AI Infrastructure
Ohio voters decided 21 data center ballot measures, at least 14 municipalities passed restrictions, and data center operators fell on November 4. What the numbers show and do not show.
Nov 5, 20268 min read
Oracle
AI Infrastructure
AI Financing
Oracle's $20 Billion Stock and Bond Raise and the $3.2 Trillion AI Obligations Question
Oracle filed to sell $10B of stock and $10B of notes for its AI buildout. We break down dilution, debt definitions, and Nikkei's $3.2T off-balance-sheet estimate.
Nov 4, 202610 min read
Revolut
GLM-5.3
Self-Hosted AI
Revolut's Self-Hosted GLM Coding Assistant: The Build vs Buy Math Behind a 90% Cost Cut
Revolut says self-hosting GLM on 56 Nvidia B200 GPUs cut AI coding costs about 90% for 2,000 engineers. We test the build vs buy economics and the vendor-number caveats.
Nov 4, 20269 min read
Anthropic
Recursive Self-Improvement
Responsible Scaling Policy
Anthropic's S-1/A Says It May Imminently Meet Its AI R&D Threshold: What the Definition Means and How OpenAI and Google DeepMind Compare
Anthropic's amended S-1 says it may imminently meet its RSP threshold for AI-driven acceleration, though no current model is designated past it. What the definition means, how OpenAI and DeepMind compare, and what to watch.
Nov 4, 202610 min read
Anthropic
IPO
S-1
Anthropic Amended S-1: $17B Q3 Revenue and $895B in Commitments
Anthropic's amended S-1 shows Q3 revenue of $16.956B, a $23.29B net loss driven by fair-value changes, and $895.4B of compute commitments. A neutral analysis.
Nov 4, 202610 min read
Nvidia
Huawei
DeepSeek
Nvidia's Reported 51% Non-Chinese Rule, DeepSeek V5 on Huawei Ascend and the HBM Loophole
Three unconfirmed chip-access reports in one week: Nvidia's reported 51% non-Chinese cloud rule, DeepSeek V5 on 160,000 Ascend 950DT chips, and smuggled HBM. What is sourced and what is not.
Nov 4, 20269 min read
Super Micro
Earnings
AI Infrastructure
Super Micro Q1 FY2027: $19.95B Revenue, 12.7% Margin and a 44% Customer
Super Micro reported Q1 FY2027 revenue of $19.95B (+298% YoY) and raised its FY2027 outlook to $72-82B. A neutral look at margins, customer concentration and what guidance implies.
Nov 4, 202610 min read
AI and Jobs
Canva
Fiserv
Canva's 1,271 Job Cuts and Fiserv's 3,000: What the Disclosures Say About AI and Restructuring
Canva affects 1,271 roles (22% of staff) and Fiserv reports 3,000 positions cut as Project Elevate doubles to $1B. What each company said about AI, and what it did not.
Nov 4, 20268 min read
Palantir
Earnings
Enterprise AI
Palantir Q3 2026 Earnings: US Commercial Overtakes Government
Palantir's Q3 2026 revenue hit $2.313B (+96%), and US commercial ($984M) passed US government ($953M) for the first time. Growth durability, Rule of 40 and bookings, explained.
Nov 4, 20269 min read
OpenAI
Model Spec
AI Alignment
OpenAI Rewrites Its Model Spec Around Five Values: What Changed and What It Means
OpenAI reorganized its Model Spec on October 29, 2026 around five values, dropped most style rules and renamed the assistant to GPT. A neutral breakdown of the changes, the training caveat and how it compares with Anthropic's approach.
Nov 2, 20267 min read
OpenAI
Google
TPU
OpenAI Reportedly Plans to Buy Google TPUs: What a Nvidia Customer Diversifying Would Mean
The Information reports OpenAI plans to buy Google's TPU chips. A neutral analysis of what is and is not known, the timeline of OpenAI's non-Nvidia compute moves, and the questions the report leaves open.
Nov 2, 20266 min read
OpenAI
Funding
Valuation
OpenAI Trims Its Funding Round to $20 Billion: What the Lower Target Says About AI Capital Markets
The Financial Times reports OpenAI cut its round from $30 billion at $1.4 trillion to about $20 billion at $1.25 trillion, with $11 billion secured so far. A neutral analysis of the numbers, the market backdrop, and what it means for the Anthropic IPO.
Nov 1, 20267 min read
AI Security
Claude Code
Cursor
A Trojanized Claude Code Extension Reached Up to 5,000 Cursor Users: How the Attack Worked and What Developers Should Do
A malicious version of the Claude Code extension, published under Anthropic's ID in Cursor's marketplace on October 29, 2026, installed an SSH key and persistence. A neutral analysis of the mechanism, the 12-hour window, who was affected, and the remediation steps.
Nov 1, 20268 min read
OpenAI
Biology
Drug Discovery
GPT-Rosalind-6: OpenAI's Biology Model Doubles Some Benchmark Scores Behind a Trusted-Access Gate
OpenAI released GPT-Rosalind-6 on October 30, 2026 for qualified life-science teams only. A neutral analysis of the benchmark gains over 5.5, the preparedness evaluations, the relaxed dual-use refusal policy, and the safety measures that got worse.
Nov 1, 20269 min read
AI Policy
Web Crawling
Publishers
One robots.txt Token for AI Training: What OpenAI, Google, Microsoft and Anthropic Agreed To, and What It Does Not Do
On October 30, 2026 four AI labs said they will honor a single robots.txt token to opt content out of training. A neutral analysis of the draft standard, rollout dates from Anthropic to Google, and its limits: training only, voluntary, and not retroactive.
Nov 1, 20267 min read
Microsoft
Azure
AI Capex
Microsoft Q1 FY2027: Azure Grows 48 Percent, and a Lease Accounting Change Reshapes the Capex Picture
Microsoft reported fiscal Q1 2027 on October 28: revenue of $92.3 billion, Azure up 48 percent to $33.2 billion, capex of $45.2 billion against an outlook above $50 billion, and a shift toward operating leases. A neutral analysis of what the numbers do and do not show.
Oct 31, 20269 min read
AI Regulation
New York
Chatbots
New York Signs an AI Chatbot Law for Minors: What S.9051-B Requires and Why Age Checks Are the Hard Part
Governor Hochul signed S.9051-B on October 28, 2026, barring chatbots from pretending to be human, claiming relationships or promoting self-harm with minors. A neutral analysis of the six feature groups, the age-assurance rule, the $25,000 penalty and the compliance questions.
Oct 31, 20268 min read
Gemini 4 Carbon
Google
Cybersecurity
Gemini 4 Carbon: Google's Security-Focused Model, Benchmarks, Pricing and Restricted Access
Google introduced Gemini 4 Carbon on October 28, 2026, scoring 88.2% on SEC-bench Pro and 54.8% on ExploitGym, with staged access through the Fairwind Program. A look at the benchmarks, the $2/$10 introductory pricing and what a restricted release means.
Oct 30, 20268 min read
Showing 51-75 of 769 posts
Experience AI That Works for You
Join thousands of users leveraging metir to boost their productivity with the world's best AI models.
Get Started Free