Agent Infrastructure
Payments, protocols, runtimes, and the cost of reasoning.
Agent InfrastructureAnalysis
AI Agents Now Clock a 16-Hour Workday
METR's benchmark shows the clean-task horizon doubling fast, but the same agents still trip on messy real work in minutes.
Agent InfrastructureRoundup
Agents Escape, Nvidia Charges, Elvis Sues
OpenAI's own agents found real ways out of their sandboxes, Nvidia priced agent security at $2,000 a chip, and Elvis Presley's estate sued Anthropic.
Agent InfrastructureAnalysis
Linux Maintainers Scrap Their AI Disclosure Tag
Linux's maintainers dropped the rule asking coders to flag AI-assisted patches, and Torvalds says what matters is that coders understand what they submit.
Agent InfrastructureNews
Anthropic Launches Claude Duet, an Always-On Agent
Claude Duet drafts replies and runs tasks on its own, answering rivals from OpenAI and Meta.
Agent InfrastructureNews
DeepMind's Agents Beat Minecraft After 605 Deaths
Four Gemini 4 Argon agents finished Minecraft on their own, after 605 deaths and 2,783 self-written programs in one 27-hour run.
Agent InfrastructureAnalysis
Claude Spends Most of Its Time Waiting on You
Anthropic's own data says Claude agents spend most of their work time waiting for a person.
Agent InfrastructureAnalysis
OpenAI Promises Codex Reset as Claude Pro Users Hit Limits
OpenAI's Codex lead pledged a daily fix or reset for 28 days, as Anthropic's own increase to Claude Pro nets out as a cut.
Agent InfrastructureAnalysis
Five Memory Vendors Claim the Lead on One Benchmark
past.dev posted the highest BEAM scores yet, four rivals claimed the same crown since April, and AI Lately grades nine memory layers.
Agent InfrastructureAnalysis, Greg Isenberg's Takes, Graded
Data Backs 4 of Greg Isenberg's 18 New Takes
Part two: Greg Isenberg posted 18 more San Francisco takes on Oct. 15, and published data backs four of them as hard trends.
Agent InfrastructureAnalysis, Greg Isenberg's Takes, Graded
Data Backs 13 of Greg Isenberg's 21 AI Takes
AI Lately graded Greg Isenberg's 21 San Francisco observations as hard or soft trends, each backed by three statistics and a chart.
Agent InfrastructureAnalysis
Jonathan Sullivan Tests Dot and Muse on Everyday Errands
Jonathan Sullivan gave OpenAI's Dot and Meta's Muse the same errands and found that memory and calendar access set them apart.
Agent InfrastructureAnalysis
OpenAI Launches Dots, Agents That Start Work Before You Ask
OpenAI unveiled dots at DevDay on Tuesday: always-on agents with their own cloud computers that read your connected apps and act on what they find. Pro plans from $100 a month and Business Premium get them first, with Pro held back in the EEA, Switzerland and the UK.
Agent InfrastructureOpinion
The Invoice Agentforce Left Uncapped
Salesforce's Agentforce revenue grew 240 percent this year by billing enterprises for units of agent work in place of human seats, and Gartner says that meter keeps climbing.
Agent InfrastructureRoundup
The Signal Brief: Sunday's Silence, Silicon, and Sovereign Capital
OpenAI's four-month silence about its own rogue agents, a $60 billion Qualcomm-Amazon chip pact, and Cohere's bid for a sovereign AI war chest show capital and disclosure moving in opposite directions.
Agent InfrastructureRoundup
The Signal Brief: Wednesday's Reckonings, Raises, and Robots
A pretraining researcher's exit, a credit fight over a machine-solved Millennium Prize problem, and Cognition's forty-eight-billion-dollar valuation show capital and conscience pulling AI in opposite directions this Wednesday.
Agent InfrastructureAnalysis, 2026 in Stories
Agent Security Becomes a Board Matter
OpenAI's agents breached Hugging Face through reward hacking, Nvidia bought the compromised platform days later, and CrowdStrike posted its best quarter ever selling defenses against exactly that scenario.
Agent InfrastructureFeature
Reasoning's Rebate: BRAID and the Economics of Bounded Reasoning
Armağan Amcalar and Eyup Cinar's BRAID framework treats AI reasoning as a bounded, auditable graph, and OpenServ's benchmark claims against GPT models are pushing enterprises to price reasoning by the dollar saved per correct answer.
Agent InfrastructureAnalysis, 2026 in Stories
Inference Economics: Gartner's 96 Percent
Gartner projects AI-optimized cloud infrastructure spending to nearly double in 2026 even as per-token prices keep collapsing, a paradox that inference specialists and reasoning researchers are racing to resolve.
Agent InfrastructureAnalysis
The Interoperability Layer: MCP and A2A
Anthropic's Model Context Protocol and Google's Agent2Agent protocol left single-company control for the Linux Foundation within roughly a year, assembling every major AI lab under one governance roof.
Agent InfrastructureAnalysis, 2026 in Stories
Machines With Money: Agent Payments Arrive
Mastercard, Visa, Stripe, Google, and Coinbase built five competing rails for agent payments inside fifteen months, and their overlapping partner rosters reveal a single converging standard taking shape.
Agent InfrastructureAnalysis
The Agent Runtime Wars
OpenAI, Anthropic, Google, Microsoft, and the open-source LangChain ecosystem each shipped a competing agent-runtime abstraction within one year, leaving orchestration itself, ahead of the underlying protocol, as agentic AI's real lock-in layer.
